Files
LocalAI/swagger/docs.go
T
mudler-agentandEttore Di Giacinto 2cadf319a0 feat(realtime): LFM3-Audio speech-to-speech on llama.cpp, remove liquid-audio (#12958)
* chore(backends): remove the liquid-audio Python backend

The backend targets liquid-audio, which loads LFM2 checkpoints only.
Remove its directory, the gallery file and the LFM2.5-Audio model
entries, the library and generated backend gallery entries, the
backend matrix and CI job, the Makefile targets and the golden test
fixtures that listed it.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* refactor(config): drop liquid-audio from capabilities, importers and UI

With the liquid-audio backend gone, remove its capability entry, its
importer, its artifact-gate entry and the realtime_audio heuristic
that only matched that backend. realtime_audio is now set only
through known_usecases. Update the Talk and fine-tuning pages, the
docs and the swagger text that named the backend.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* chore(llama-cpp): bump llama.cpp to e486a4b5

Move the llama.cpp pin to e486a4b559617dc88bf79978996f6c670b6bbc9d,
the master head of 2026-10-28. It contains the LFM3-Audio support
(ggml-org/llama.cpp#31257) and the realtime server split
(ggml-org/llama.cpp#31269).

The score task patch is regenerated against the new pin. It applies
with git apply and with patch -p1, with no offset and no fuzz. Its
hunks in server-task.h and server-context.cpp were rebased by hand
next to SERVER_TASK_TYPE_REALTIME and the control dispatch.

The C++ backend was built on CPU for this change: the llama, mtmd and
server-context libraries and the grpc-server binary link.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(llama-cpp): compile server-realtime.cpp into grpc-server

server-context.cpp includes server-realtime.h, and its realtime slot
type is defined in server-realtime.cpp. Upstream compiles that file
into the server library. The gRPC server is one translation unit, so
it must include the file too, or the link fails.

The include is guarded by __has_include. Forks that predate the
realtime split do not ship the file, so they keep building as before.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* chore(llama-cpp): probe realtime and TTS voice support in prepare.sh

prepare.sh writes llama_compat.h from probes on the checkout. Add two:
LOCALAI_HAS_REALTIME (SERVER_TASK_TYPE_REALTIME in server-task.h) and
LOCALAI_HAS_TTS_VOICE (set_voice in mtmd-helper.h). The probes let the
gRPC server choose the realtime and voice code paths per checkout, so
forks without them keep compiling.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(llama-cpp): advertise speech recognition and speech-to-speech

llama-cpp now registers the AudioTranscription and AudioToAudioStream
RPCs, with the transcript and realtime_audio usecases. The default
usecase stays chat, so a plain GGUF is still a chat model. The
backend implements the RPCs in a follow-up change; until then the
calls return UNIMPLEMENTED.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(config): add pipeline.any_to_any for speech-to-speech models

A realtime_audio model that serves the whole speech loop itself sets
pipeline.any_to_any. The flag is off by default, so chained pipelines
and existing realtime_audio configs are unchanged. The realtime_audio
heuristic stays explicit: a llama-cpp model is never a Talk-page model
unless known_usecases says so.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(config): read the audio encoder and speech-to-speech projector flags

Add HasAudioEncoderProjector for clip.has_audio_encoder and
HasSpeechToSpeechProjector, which requires the LFM3 audio encoder
family together with the gen-audio decoder. Qwen3-TTS sets both encoder
flags too, so the flags alone cannot tell it apart from LFM3.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(importer): import LFM3-Audio projectors as speech-to-speech

An mmproj that carries the LFM3 audio encoder and the gen-audio decoder
now imports with known_usecases realtime_audio, chat, transcript and
tts, pipeline.any_to_any: true, the parallel:2 option and the tokenizer
template. Other gen-audio projectors keep the TTS-only result, Qwen3-TTS
included. The probe runs against a local HTTP server in the tests.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(llama-cpp): named-voice TTS and the LFM3 ASR prompt

Models without a speaker reference (LFM3-Audio) take a named voice
through their "tts" chat template instead of a reference file. The
request is validated with needs_reference=false, the prompt is rendered
with common_chat_get_tts_prompt(), and the voice is set on the gen-audio
input. Qwen3-TTS Base keeps the reference-file rule. The branch is
guarded by LOCALAI_HAS_TTS_VOICE, so forks without named voices keep the
old behaviour.

Transcription takes its prompt from common_chat_get_asr_prompt(). LFM3
gets a system turn that names the language and an empty user turn.
Qwen3-ASR's preset is unchanged: fixed user text, no system turn, and a
caller prompt is still appended after a newline. The preset call is
guarded by a new LOCALAI_HAS_ASR_PRESET probe.

The pure parts sit in tts_request_options.h and transcription_prompt.h,
with standalone tests run by backend/cpp/run-unit-tests.sh.

Add the test-extra-backend-llama-cpp-lfm3 target. It runs the tts,
ttsstream and transcription specs against LFM3-Audio GGUF weights. It is
a manual run and is not wired into CI.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): mock AudioToAudioStream and pending realtime S2S specs

The mock backend serves AudioToAudioStream deterministically: session
created, commit with RMS, a tool call on the first turn with tools, then
transcript pieces and paced 24 kHz tone frames, with response.cancel
ending a reply as cancelled.

The realtime S2S specs cover the mock and real-weights variants. They
skip until the /v1/realtime S2S path lands, so the mock suite stays
green in CI.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(llama-cpp): speech-to-speech AudioToAudioStream on the realtime slot

Implement AudioToAudioStream on upstream's realtime slot, so a model
that answers after the audio (LFM3-Audio) can hold a conversation over
the existing duplex RPC. The proto schema does not change: tools travel
in the config, tool calls leave as function_call_arguments.done and tool
results arrive as a function_call_output item.

realtime_session.h keeps what needs no model: the turn text derived
from the chat template (with a tool turn), the conversation state, a
filter that keeps <tool_call> blocks out of the transcript and drops
audio produced inside them, and the translation of slot results into
events. grpc-server.cpp wires it to the slot with control tasks and
parses calls with the parser the chat template selects.

The unit test uses a fake renderer. A ctest-only check renders the real
LFM3 template and compares the slot text with the full render.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(gallery): add the LFM3-Audio-2B entries

Two llama-cpp entries for LiquidAI/LFM3-Audio-2B-GGUF, the Q4_K_M
default (lfm3-audio-2b) and the Q8_0 build (lfm3-audio-2b-q8), both
with the Q8_0 audio projector. Each declares pipeline.any_to_any and the
realtime_audio, chat, transcript and tts usecases the importer writes
for an LFM3 projector, with parallel:2 and the tokenizer template. The
Q4 entry lists the Q8 one as a variant.

File sizes and sha256 come from the Hugging Face API for the pinned
repo revision. Speaker reference files and the VAD model are not part
of the entries: VAD stays optional.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* docs: describe LFM3-Audio speech-to-speech and its gallery entries

Adds a speech-to-speech section to the Realtime API page (any_to_any
configuration, VAD versus manual commit, tools, transcripts, barge-in,
limits), a pipeline.any_to_any row, a llama.cpp speech-to-speech row in
the compatibility table, and LFM3-Audio mentions on the TTS and ASR
pages. The feature is marked experimental: the behaviour follows the
backend session contract and has not been run with real weights.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(realtime): run any-to-any models on AudioToAudioStream

A realtime_audio model with pipeline.any_to_any now opens one duplex
stream to the backend and keeps it for the session, replacing the
"any-to-any models not implemented" branch. The chained pipeline is
unchanged.

The endpoint forwards input audio (with a pre-roll from speech start
when a VAD model is set, otherwise every chunk and a client commit),
sends response.create, and turns the backend's events into the usual
realtime events: audio and transcript deltas, tool calls through the
existing function_call items, and a single response.done. A function
call output is forwarded before the next response. Barge-in sends
response.cancel ahead of the new speech and stops forwarding audio at
once. The user's speech is transcribed on the side and reported on the
user item; a failure there does not touch the response.

Server VAD is on only when pipeline.vad is set. Classifier mode,
compaction, streaming.llm, voice gating, sound detection and
semantic_vad are rejected for these models.

The mock backend now matches the llama.cpp contract (session.updated,
the cancel error when idle) and its VAD reports the loud stretches of
the audio, so the S2S e2e specs run a VAD turn. The specs are enabled.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): add the real-weights realtime S2S variant and the gRPC s2s spec

The realtime S2S specs now check reply audio on real models: duration,
loudness, clipping, and keywords found in the model transcript or in a
separate ASR of the output audio. run-realtime-lfm3.sh stages the llama-cpp
and sherpa-onnx backends, the LFM3-Audio files and the sherpa VAD, ASR
and TTS models, then runs the real variant. It is manual and never part of
CI: `make test-realtime` still filters out real-models.

The backend suite gains an s2s capability (spoken turn, tool call, tool
result, spoken answer over AudioToAudioStream) and optional TTS duration
and RMS bounds plus a transcript keyword check, all enabled for
test-extra-backend-llama-cpp-lfm3.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): expect the idle-cancel error from the mock S2S

The mock now answers a response.cancel with no reply running with
response_cancel_not_active, as the llama.cpp backend does. The cancel
spec still treated that cancel as a no-op and read the error as extra
data after the stream ended.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(talk): show any-to-any models as one speech-to-speech model

/api/pipeline-models lists a pipeline.any_to_any model with any_to_any
set and its optional VAD. A realtime_audio model that is not any-to-any
is listed only when it has the four pipeline slots, since the session
takes the chained path for it.

The Talk page shows a speech-to-speech model as one model: no four-slot
view, the VAD slot only when a VAD model is set, and the female and male
voices offered. Without a VAD model the server does not detect turns, so
the page offers Done speaking, which sends input_audio_buffer.commit and
then response.create. Text-only replies stream through
response.output_text.delta and .done.

self_contained stays in the response for clients built before any_to_any.

Assisted-by: Claude:claude-haiku-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): refuse a llama-cpp realtime model without pipeline.any_to_any

llama.cpp has no VAD, transcription or TTS RPCs, so filling the empty
pipeline slots with the model name failed later with UNIMPLEMENTED.
Fail at session start and name the missing setting instead.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): hold manual-commit S2S audio until the commit

In manual-commit mode the endpoint streamed every appended chunk to the
backend, so input_audio_buffer.clear only emptied the Go buffer while the
model kept the audio, and idle audio filled its context between turns.
Keep the audio in the input buffer and send it, behind pending text and
tool-result items, when the client commits or asks for a response.

Frames and controls now go through one ordered write queue, so no gRPC
send happens under the audio buffer lock. A tool result created while a
VAD turn is open is refused with a clear error. When the stream ends
outside a response the session reports it and closes, and a bare EOF
from Send reports the status that Recv holds.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): drop the S2S readiness gate and stale comments

The gate was always true. Some comments referred to work items that
readers of the repository cannot see.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* docs(realtime): add the liquid-audio migration note and the clear semantics

Say that the Python liquid-audio backend is gone, which gallery entries
replace it and that LFM2.5-Audio has no replacement. Describe what
input_audio_buffer.clear drops for speech-to-speech sessions now that
manual-commit audio is held back until the commit.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(talk): do not promise turn detection when no VAD model is set

The idle note said the server notices when you stop talking. A
speech-to-speech model without a VAD model needs Done speaking, so the
note now says that.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): make the mock S2S backend keep the llama-cpp session limits

The mock accepted what the real backend refuses and hid two endpoint bugs.
It now treats output_modalities from the config as a fixed session mode and
picks the default per response from the tools set at that time. A fixed text
plus audio mode with tools reads the tool call out loud. After the first
commit or message, session.update of instructions, voice or tools is an
error that does not belong to a response, committing an empty buffer is an
error, and response.create while a reply runs is refused. Errors of a
refused response.create carry response_scoped in their meta.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): leave the S2S output mode to the backend unless the client sets one

The stream config sent text plus audio when the session had no tools yet.
The llama-cpp backend keeps a mode from the config as a fixed session mode,
and tools added later by session.update never cleared it. The model then
read tool calls out loud. Send output_modalities only when the client chose
them. The backend picks the default of each response from the tools it has
when the response starts.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(llama-cpp): mark the realtime errors that answer a response.create

The error event now carries response_scoped: true when it ends the request
for a response (a refused response.create, a failed commit made by it, or
a stream failure). Other errors, such as a refused session.update or a
commit of an empty buffer, are not about a response and one that runs
keeps generating. The client needs the difference to know whether a
response.done is still coming.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): end an S2S response only on the errors that belong to it

Every backend error ended the response in flight, although the backend
also reports errors about other things (a refused session.update, a commit
of an empty buffer) and goes on generating. The rest of the reply was lost
and the next response.create met either a refusal or the leftover events of
the old reply. End the response only on an error marked response_scoped,
forward the others to the client and keep waiting for response.done. Drop
the events buffered by an earlier response when a new one starts, and skip
the commit of a turn that forwarded no audio.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): run the S2S input transcription detached from the commit slot

The commit waited for the side transcription of the user audio while it
held the commit slot, and response.create waits for that slot. The ASR
latency was added before every reply, and on a backend with few slots a
busy ASR task could stall the reply. The model already has the audio, so
run the transcription detached under the session context and release the
slot at once. The transcript is still reported on the user item.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(realtime): do not pin the place of the input transcript in the S2S order

The transcript of the user audio is reported on the side and may land
before or after the first response events. Check the order of the response
events without it.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): stop an S2S reader leak and a cancel that waits on the write queue

The reader waited on a channel it closes itself when the events buffer was
full, so it outlived the session after a turn ended on an error or timed
out. It now also stops when the session closes.

Stopping a response sent the cancel through the write queue and waited for
it before the grace timer started. With backpressure on the stream the wait
never ended and nothing read the events meanwhile. The timer now starts
first and the cancel is only queued.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(realtime): let a chained pipeline run without a VAD model

A pipeline that names no pipeline.vad failed to build, because the
VAD slot was always loaded, and a new session still started with
server_vad. The slot is now optional for chained pipelines as it
already is for any-to-any models: the session starts with turn
detection off and the client commits each turn. Turning server_vad on
for such a session is refused, and a switch to a pipeline without a
VAD drops turn detection.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(gallery): add LFM3-Audio VAD and chained-pipeline entries

Add lfm3-audio-2b-vad and -q8-vad (any-to-any with silero-vad-ggml as
the turn detector), and lfm3-audio-2b-pipeline, -pipeline-vad and the
-q8 forms (the model chained through its own transcription, chat and
speech slots, with and without the VAD). The gallery has no
dependencies between models, so the VAD entries state that
silero-vad-ggml must be installed first.

Drop the variants list from lfm3-audio-2b: it let the installer pick
the larger Q8 build on any host with enough memory, so a plain install
of the Q4 entry did not get Q4. All entries are now independent. Tests
cover the entry shapes, the single-build install and their acceptance
by the realtime endpoint.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): run the realtime specs for any-to-any and chained shapes

The realtime specs covered one any-to-any model with a VAD. They now run
for any-to-any and for a chained pipeline, each with and without a VAD,
on the mock backend in CI. VAD-less shapes also check that turn
detection starts off and that server_vad is refused. The manual
real-weights script writes one config per shape that mirrors the gallery
entries and runs the four shapes in turn (LFM3_MATRIX picks a subset).

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* docs(realtime): document the LFM3-Audio VAD and chained-pipeline entries

List the eight LFM3-Audio gallery entries, explain that silero-vad-ggml
must be installed before the -vad entries, describe the chained pipeline
on one model and the behaviour of a pipeline without a VAD, and
describe the shape matrix of the e2e specs.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(realtime): wait for the detached input transcription in the S2S specs

The input transcription runs in its own goroutine, so the transcription
completed and failed events can arrive after the response ends. Poll for
them instead of reading the event list right after the response.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): release the S2S stream before running server-side tool calls

A response held the stream slot until it returned, and it ran the tool
calls from inside that scope. A server-side tool starts a follow-up
response on the same stream, which then waited for the slot its parent
still held, so the client never got response.done. Run the tool calls
once the slot is free; the event order is unchanged.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): end an S2S session whose stream dies, and stop queuing after close

A stream that died during a response left the session open: later
requests failed one by one and VAD frames kept being forwarded. The
response now ends the session through the transport, and enqueue drops
requests once the session is closed so the write queue cannot grow.

The writer is now the only caller of Send and of CloseSend. close()
used to call CloseSend while the writer could be inside Send, which
gRPC does not allow.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): close the open S2S turn when the VAD stops

Switching turn detection to manual in the middle of a VAD turn left the
turn forwarding, and the next client commit sent the whole buffer a
second time. Stopping the VAD now commits the forwarded audio on the
backend and drops it from the input buffer.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): keep a tool result pending while an S2S user turn is open

The check that refuses a tool result during a user turn ran when the
item arrived, but the item was forwarded later, and the result was
marked as sent before the backend took it. If the user spoke in
between, the backend refused it and it was never retried. The result is
now claimed and queued under the audio lock only when no turn is open,
and otherwise stays pending for the next commit or response.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(talk): empty the input buffer for models without a VAD

Without voice detection the microphone fills the server's input buffer
for the whole session, so Done speaking committed idle silence and the
echo of the previous reply. Clear the buffer when the session starts and
when a reply ends.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(importer): keep an existing parallel option on LFM3-Audio imports

The speech-to-speech import added parallel:2 even when the options
already carried another parallel value. Use the same check as the
hardware defaults, which also covers n_parallel.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(talk): list chained pipelines that have no VAD

/api/pipeline-models required a VAD for a chained pipeline, so the
VAD-less LFM3-Audio pipeline entries never showed up on the Talk page.
Any pipeline with transcription, LLM and speech slots now qualifies.
The Talk page treats every model without a VAD as manual-turn, and the
session sheet says so for chained pipelines too.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* chore(llama-cpp): bump llama.cpp to 0ac1a7fd

Move the llama.cpp pin to 0ac1a7fd35c127a18c0cbc6c57d12e256a6a385b,
the master head of 2026-10-29. It contains tool calls in the realtime
server (ggml-org/llama.cpp#31287).

The score task patch is regenerated so that it applies with no offset
and no fuzz.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(functions): keep a null tool argument null

A tagged parameter whose text was "null" came out as an empty string:
unmarshalling JSON null into a string succeeds and leaves it empty, so
the value was taken for a string. Check the first byte of the JSON
value instead. A parameter with the text "null" now becomes JSON null,
as in llama.cpp.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(functions): add an lfm3 XML format preset

The LFM3 chat template writes a call as <tool_call>, <function=NAME>,
one <parameter=KEY> block per argument, then the closing tags. The
qwen3-coder format has the same shape and auto-detection already found
it, but xml_format_preset: lfm3 was not a known name.

Add the preset, with tests for one call, a call without arguments, two
calls, string and non-string values (numbers, booleans, null, arrays,
objects), a call between text, and unfinished input.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(realtime): parse tool calls with the chat template, as upstream

Since ggml-org/llama.cpp#31287 the realtime server handles tools with
the parser and grammar of the chat template. Do the same in the
speech-to-speech stream instead of a hard-coded tag filter and a parse
of the whole text at the end.

- Each response gets a common_chat_session with the tools of the
  session and parallel calls. Its grammar and preserved tokens go into
  the sampling of the response, so the slot keeps the call tokens in
  the text and constrains the call.
- The text of the response is fed to the session piece by piece
  (task_result_state::update_chat_msg). Content deltas become the
  transcript; tool-call deltas are collected into one
  function_call_arguments.done per call at the end. Audio after the
  start of a call is dropped.
- A tool_choice of auto, none or required is accepted in the params of
  the config, in session.update and in response.create, with the
  response overriding the session.
- The tool turn is derived from a conversation in which the assistant
  called a tool, like upstream, not from a plain reply.
- response.done carries raw_text, everything the model wrote with the
  call markup, for a caller that parses it another way. If the parser
  throws, the rest of the response is content and the raw text is the
  way out.

ToolCallFilter is no longer used and goes in the next commit.

The unit tests cover the splitter contract (content, call deltas, the
final piece, whole characters only). realtime_chat_check feeds
LFM3-format text through the session in pieces of 1, 7 and all bytes
and checks the call, the transcript and the grammar.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* refactor(realtime): drop the tool-call text filter

The parser of the chat template now splits content from tool calls, so
ToolCallFilter, which looked for the literal <tool_call> tags, has no
caller. Remove it with its tests. Every case it covered (a call
between text, split tags, several calls, an unfinished call, text after
a call) is handled by the parser, and realtime_chat_check runs the
LFM3 format through it in pieces of different sizes.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(realtime): fall back to the raw text for S2S tool calls

For an any-to-any model, the calls the backend parsed still win. When
there are none and the request had tools, parse the raw text that
response.done now carries the way the chained pipeline parses the text
of its LLM (CleanupLLMResult, ParseFunctionCall, ParseTextContent), so
a call format the backend parser does not know can still be served
with function.xml_format_preset or the other function settings of the
model. The content around the calls replaces the streamed transcript
only if the config captures it, as in the chained path.

Also pass tool_choice to the backend: the session mode with the stream
config and with session.update, and the mode of a response.create
override with that event. A named function cannot be forced, so it is
not sent.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): carry raw_text in the mock S2S response.done

Match the llama-cpp session: with tools set, response.done has raw_text,
the call markup of the LFM3 template for a tool turn and the transcript
for any other reply.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(functions): parse tool blocks that have text between them

The iterative XML parser stopped at the first scope block that was
preceded by text, so "<tool_call>A</tool_call> then <tool_call>B" gave
one call. Skip text between blocks once a call was found; text before
the first block still ends the parse as before. Blocks that follow each
other, or only whitespace, already worked.

Tests cover three calls back to back and with text between them, with
the lfm3 preset and with auto-detection. They also pin the current
reading of argument values: the parser has no tool schema, so a value
such as 42 or true is read as JSON even when the schema types it as a
string.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(gallery): read LFM3-Audio tool calls with the lfm3 preset

Set function.xml_format_preset: lfm3 on the eight LFM3-Audio-2B entries
and in the config the importer writes for a speech-to-speech projector,
so the server parses the call format of the LFM3 chat template when the
backend returns no call (and in the chained pipeline entries, which use
the regular chat path). gallery/index.yaml is regenerated from the
library.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* docs(realtime): describe how S2S tool calls are read

Replace the description of the hard-coded tag filter with the parser
and grammar of the chat template, tool_choice, and the raw-text
fallback in the server. Document function.xml_format_preset, and set it
in the real-weights e2e model configs like the gallery entries do.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(functions): keep earlier calls when a later block does not parse

After a valid call, text that holds the opening tag again (a prose
mention, a markdown example, a malformed second block) made the parser
treat it as a new block, fail, and return no calls at all. Remember the
state after each complete block and fall back to it when a later one
fails: the calls found so far are kept, parsing stops there and the
remaining text stays in the content.

Text skipped between two valid blocks now goes to the content instead of
being dropped, so StripToolCallMarkup keeps it.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): do not drop audio after a tool call started

In mixed mode the speech for text written before a call can still be
pending when the call starts, so dropping every audio frame after the
first call delta cut the speech while its transcript was sent. The
llama.cpp server never drops audio when it reads a result; do the same.
If a model speaks the call markup aloud, that is model behaviour to
observe on real weights, not something to filter here.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): flush held-back text when a response is cancelled

The parser got is_partial=true on a cancelled response, so text it held
back (a possible start of a tag) never reached the transcript. The llama.cpp
server parses a cancelled response with is_partial=false; do the same. A
cancelled response still makes no tool call. The path where the parser
throws already sends the text as content and needs no change.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(realtime): stop the fake S2S stream from sending after Close

The barge-in spec failed now and then with a data race and "send on
closed channel": its script pushes audio from a goroutine, and the
session closes the stream while that goroutine still runs. A push now
does nothing once the stream is closed, and Close waits for a push in
flight. The production close order is right: the writer closes the
stream only after it stops sending.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): align the S2S tool-call fallback with the backend

Four gaps in how the Go side reads the tool calls of an any-to-any reply:

- With tool_choice none (the response value wins over the session value)
  the backend builds no tool parser, yet the raw-text fallback still
  turned a call the model wrote into a function_call item. Skip it.
- Without a capture regex the fallback never cleaned the transcript, so the
  stored item and transcript.done kept the raw call markup. When the call
  is XML, use the text left after the markup is cut out; a reply that is
  only a call now has an empty transcript.
- A named function as tool_choice mapped to nothing, so the backend kept
  the previous mode. Refuse it with an error in session.update and
  response.create.
- A response-level tools list changed what the fallback looked for, but
  the backend only has the tools of the session. Use those in both.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* docs(realtime): say how S2S tool_choice and response tools behave

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(e2e): use the real llama-cpp backend in the real realtime run

The suite always registered the mock backend under the name llama-cpp, so
a run with REALTIME_S2S_MODEL set never reached the llama-cpp backend
found in REALTIME_BACKENDS_PATH. Register the mock only when that run is
off.

The real run converts the input audio with ffmpeg. Check for it in the
suite and in run-realtime-lfm3.sh before any download or backend start,
with a message that says what to install, and list it as a prerequisite
in the script header and the Makefile target.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(realtime): cover the tools path of a client session.update to the S2S backend

A tool call that never came back from an any-to-any model raised the
question whether the tools of the session reach the backend at all. Add a
spec that sends the session.update of the real-weights e2e (flat
Responses API tool, no session type) and checks that the backend gets
the tool in the chat-completions form.

Log the number of tools sent when the stream opens and when a
session.update carries them, so a run log shows it without reading the
wire. The "No tools received" line of the llama-cpp backend does not
show it: the server prints it for the input transcription request, which
has no tools by design.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): interrupt the chained pipeline reply while it is in flight

With unary stages, the default of the lfm3 pipeline gallery entries, the
chained pipeline synthesizes the whole reply and sends it as one audio
delta right before response.done. The barge-in spec waited for that
delta and then rendered the interruption with the TTS, so the first
response was already done when the speech started and ended completed
instead of cancelled. Real runs on a GPU showed it with both
quantizations.

For a pipeline, start the interruption at response.created, which comes
before the LLM and TTS stages. Render the interruption audio before the
first utterance is sent, so nothing sits between the reply and the
interruption. The any-to-any shape still waits for the first audio
delta, which that backend streams while it generates.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): accept a number word and its homophone in the spoken-answer check

The check that the reply audio says what the transcript says needs a
shared word of 4+ letters. A short answer failed it: the model wrote
"Four" and the ASR heard "for". Compare numbers in one spoken form, so
digits, number words and their homophones match each other. A homophone
only counts when the other text has the number itself, so "for" on both
sides still shares nothing and the check does not pass on a stop word.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(llama-cpp): ask LFM3-Audio for function calls when a realtime session has tools

The default system prompt of a realtime session was "Respond with
interleaved text and audio." whatever the tools. Under that prompt
LFM3-Audio answers in words ("I'll check the weather ...") and never
writes the call. The model card, and the llama.cpp realtime example
client, use "Respond with function calls, and interleaved text and
audio." when the session has tools, and the model then calls the tool.

Pick the default by the tools of the session, and treat the plain default
that the LocalAI server sends back as no choice: it follows the tools too,
also when they arrive with a session.update before the first audio. A
prompt the client wrote is kept as it is.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): transcribe the input of an any-to-any turn between responses

The input transcript of a speech-to-speech turn ran at the same time as
the reply, on the same model. llama.cpp serves both from one context: the
realtime slot wants embeddings on, the chat slot of the transcript wants
them off, and each flip makes it reserve its compute buffers again. In a
GB10 run that was two reserves of about 175 ms per step, the reply fell to
2.6 tokens per second and the transcript to 380 ms per token, until the
transcript was done.

When pipeline.transcription is the model of the stream, run the transcript
after the response, or before one that is not started yet, and keep
responses out while it runs. A separate transcription model still runs in
parallel.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): send the input transcript of an any-to-any turn before its reply

The previous change moved the same-model transcript after the response.
A client that reads until response.done then never saw
conversation.item.input_audio_transcription.completed: the transcript
was made (the log shows it) but sent after the end of the window.

Run it in line with the commit instead. A VAD turn transcribes and then
responds. A client response.create already waits for the commit slot, so
it starts once the transcript is out. The transcript event now comes
before any output of the response of its turn, for a manual commit and for
a VAD turn, and never overlaps the response on the llama.cpp context. A
transcription model of its own still runs in parallel.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(realtime): transcribe the input of an any-to-any model only when asked

The stream answers from the audio, so the transcript of the input is for
the client only. On the model of the stream it is costly: llama.cpp
serves the transcript request and the realtime stream from one context
and rebuilds its compute buffers on every step while both are active.
On a GB10 that was 26 to 29 reserves of about 175 ms per turn, against 0
to 1 on the llama.cpp server, and the reply ran at a few tokens per
second.

An any-to-any model no longer fills pipeline.transcription with its own
name. Without it the session does not transcribe the input; with another
model it does, in parallel with the reply; with the model itself named it
does, before the reply. The reply transcript always comes from the stream.
The gallery entries set none, so they get the fast path.

The mock configs and run-realtime-lfm3.sh name a transcription model, so
the real-weights spec still covers the input transcript.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* fix(realtime): end a response as cancelled when a barge-in lands during the LLM call

The chained pipeline cancels the response context on a barge-in. When that
happens while the model is generating, the backend call fails with the
context error, and the response ended as failed: an error event and no
response.done. A client that waits for the terminal of the interrupted
response then waits forever. On a GB10 the q4 model answered before the
barge-in arrived, so only the slower q8 model showed it: the spec read
events for 30 s and timed out.

When the context is cancelled, a failed Predict call or a failed result is
a cancel, as the TTS step and the streamed LLM path already treat it. The
response ends with response.done{cancelled} and no error.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): wait for the input transcript that comes from a model of its own

An any-to-any model now gets its input transcript from a separate ASR
that runs next to the reply. The reply is often over first, so a spec that
stops reading at response.done saw no transcript in the VAD shape. Keep
reading, up to 30 s, until the transcript arrives. The contract is that it
comes, not when.

Add a guard spec for what the input transcription is handed: the whole
turn, in a WAV that says the rate it was recorded at (the session rate for
a client commit, 16 kHz for a VAD turn).

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): check the errors of Close in the realtime specs

golangci-lint (errcheck) flagged the unchecked Close of the websocket
connections, the response body and the fake stream. Use the
`defer func() { _ = x.Close() }()` form the other e2e specs use.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* test(e2e): write the mock S2S specs with Ginkgo and drop the env read

forbidigo only allows Ginkgo/Gomega specs and no os.Getenv outside the
sanctioned places. Convert the mock S2S tests to Describe/It with a suite
bootstrap (the package had none), with the same 11 cases. The mock no
longer reads MOCK_S2S_FRAMES: only these specs set it, so the reply length
is a package variable they change per spec and restore afterwards.

Assisted-by: Claude:claude-sonnet-5-5
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

---------

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Co-authored-by: Ettore Di Giacinto <mudler@localai.io>
2026-10-29 22:06:41 +01:00

15588 lines
584 KiB
Go
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
// Package swagger Code generated by swaggo/swag. DO NOT EDIT
package swagger
import "github.com/swaggo/swag"
const docTemplate = `{
"schemes": {{ marshal .Schemes }},
"swagger": "2.0",
"info": {
"description": "{{escape .Description}}",
"title": "{{.Title}}",
"contact": {
"name": "LocalAI",
"url": "https://localai.io"
},
"license": {
"name": "MIT",
"url": "https://raw.githubusercontent.com/mudler/LocalAI/master/LICENSE"
},
"version": "{{.Version}}"
},
"host": "{{.Host}}",
"basePath": "{{.BasePath}}",
"paths": {
"/3d/animate": {
"post": {
"tags": [
"3d"
],
"summary": "Creates a 3D animation (binary glTF / GLB).",
"parameters": [
{
"description": "Named conditioning inputs and model-specific parameters",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Model3DAnimationRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/3d/generations": {
"post": {
"tags": [
"3d"
],
"summary": "Creates a 3D asset (binary glTF / GLB) from a conditioning image.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Model3DRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/3d/remesh": {
"post": {
"consumes": [
"multipart/form-data"
],
"produces": [
"model/gltf-binary"
],
"tags": [
"3d"
],
"summary": "Applies watertight print remeshing to an existing 3D asset.",
"parameters": [
{
"type": "string",
"description": "3D model name",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "Source GLB",
"name": "mesh",
"in": "formData",
"required": true
},
{
"type": "number",
"description": "Detail size as percent of the source bounding-box diagonal (0.35–2.5; default 0.5)",
"name": "detail",
"in": "formData"
}
],
"responses": {
"200": {
"description": "Remeshed GLB",
"schema": {
"type": "file"
}
}
}
}
},
"/api/addons/apps": {
"get": {
"description": "Returns the installed addons that are enabled, for every authenticated caller who holds the addons feature. The owner decides what a caller sees; hiding an entry is never the control, because every call the entry leads to is authorized again. Each app has its title, description, icon URL, version and installed generation, a process state (running, starting, failed or stopped; the state is an observation, separate from the enabled flag), the path of its main page when its manifest declares a valid one, and a bounded tool summary. A headless addon is listed without a main page. Administrators additionally see installed addons that are disabled and, per app, the start and stop controls and the management link. An app carries update update-pending, separate from state, only when the owner's cached update check says an update is available for the installed package; the list never evaluates a check and never reaches a library.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "List the addon apps the caller may use",
"responses": {
"200": {
"description": "Apps visible to the caller",
"schema": {
"$ref": "#/definitions/apps.List"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature or owner policy denied the request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/apps/{id}/icon": {
"get": {
"description": "Streams the immutable icon image of the app at the generation named in the query. The image comes from the verified package on the owner, relayed through the receiving frontend; it is never served from an addon process and never from a URL the manifest names. It is served only for an app the caller may see at the generation that was listed. An unknown app, a hidden app, an app without a verified icon and a stale generation give the same 403. The response is an image with X-Content-Type-Options nosniff and Cache-Control private, no-store.",
"produces": [
"image/png",
"image/webp",
"image/jpeg"
],
"tags": [
"Addons"
],
"summary": "Read the icon of an addon app",
"parameters": [
{
"type": "string",
"description": "Addon ID",
"name": "id",
"in": "path",
"required": true
},
{
"type": "integer",
"description": "Installed generation from the apps list",
"name": "generation",
"in": "query",
"required": true
}
],
"responses": {
"200": {
"description": "Icon bytes",
"schema": {
"type": "string"
},
"headers": {
"Cache-Control": {
"type": "string",
"description": "private, no-store"
},
"X-Content-Type-Options": {
"type": "string",
"description": "nosniff"
}
}
},
"400": {
"description": "Missing or malformed generation",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the app is unknown, hidden, has no icon or the generation is stale",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/apps/{id}/start": {
"post": {
"description": "Administrators only. Records the desired state enabled through the same lifecycle manager and owner as the typed enable action; the owner starts the process and health stays a separate observation. The body carries the generation the caller saw in the apps list: a different installed generation is denied before anything is recorded. The operationId is the identity of the change: repeat it to replay safely, and after an outcome_unknown response poll operation-status with it instead of submitting a new operation. Non-administrators are denied (403) whatever the body says.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Start an addon app",
"parameters": [
{
"type": "string",
"description": "Addon ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation and the generation the caller saw",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/apps.StartRequest"
}
}
],
"responses": {
"200": {
"description": "Operation accepted; the payload is the operation status",
"schema": {
"$ref": "#/definitions/contract.Result"
}
},
"400": {
"description": "Malformed body, missing operationId or generation",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or generation denied the request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable, or the outcome is unknown (type outcome_unknown, with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/apps/{id}/stop": {
"post": {
"description": "Administrators only. Records the desired state disabled through the same lifecycle manager and owner as the typed disable action; the owner drains and stops the process and keeps all data. The body carries the generation the caller saw in the apps list: a different installed generation is denied before anything is recorded. The operationId is the identity of the change: repeat it to replay safely, and after an outcome_unknown response poll operation-status with it instead of submitting a new operation. Non-administrators are denied (403) whatever the body says.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Stop an addon app",
"parameters": [
{
"type": "string",
"description": "Addon ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation and the generation the caller saw",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/apps.StartRequest"
}
}
],
"responses": {
"200": {
"description": "Operation accepted; the payload is the operation status",
"schema": {
"$ref": "#/definitions/contract.Result"
}
},
"400": {
"description": "Malformed body, missing operationId or generation",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or generation denied the request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable, or the outcome is unknown (type outcome_unknown, with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/artifacts/{id}": {
"get": {
"description": "Returns the caller's own artifact: its state (transferring, complete, failed or cancelled), the sanitized file name, the media type, the exact size and the expiry. Only the principal that requested the result can read it; an unknown identifier and another user's artifact give the same 403, so identifiers cannot be probed. The response never carries a storage path or an addon handle. This is the status lookup of the background transfer; it is read-only.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Read the state of a private addon file result",
"parameters": [
{
"type": "string",
"description": "Artifact ID (host generated, opaque)",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Artifact state",
"schema": {
"$ref": "#/definitions/gateway.Artifact"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the artifact is unknown, expired or not the caller's",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/artifacts/{id}/download": {
"get": {
"description": "Streams the caller's own complete artifact as an attachment, relayed through the receiving frontend from the owner that holds the bytes. Access needs the session or key of the principal that requested the result and, at the start and while streaming, current read access to every source resource and a current account. An artifact that is not complete, is expired, belongs to another principal or is unknown gives the same 403. The response is always an attachment (RFC 6266) with a sanitized name, a media type from a fixed allowlist or application/octet-stream, X-Content-Type-Options nosniff and Cache-Control private, no-store; it is never rendered inline. A failure after the first byte aborts the connection, so a truncated file never looks complete. Disabling the addon denies downloads and keeps the bytes until expiry; enabling it again serves unexpired artifacts to principals that still have access.",
"produces": [
"application/octet-stream",
"text/csv",
"application/json"
],
"tags": [
"Addons"
],
"summary": "Download a private addon file result",
"parameters": [
{
"type": "string",
"description": "Artifact ID (host generated, opaque)",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Artifact bytes",
"schema": {
"type": "string"
},
"headers": {
"Cache-Control": {
"type": "string",
"description": "private, no-store"
},
"Content-Disposition": {
"type": "string",
"description": "attachment with a sanitized filename and filename*"
},
"Content-Length": {
"type": "string",
"description": "exact size of the artifact"
},
"X-Content-Type-Options": {
"type": "string",
"description": "nosniff"
}
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the artifact is unknown, expired, incomplete, not the caller's or no longer readable by the caller",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable, or the download bound is reached",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/builtin": {
"get": {
"description": "Returns the page add-ons that ship inside the web bundle. The list is empty today: Studio and Talk are packaged add-ons that a person installs from the Add-ons drawer. Each entry carries only its own gate: the route needs an authenticated caller and no ` + "`" + `addons` + "`" + ` feature.",
"produces": [
"application/json"
],
"tags": [
"AddonNavigation"
],
"summary": "List the built-in add-ons the caller may open",
"responses": {
"200": {
"description": "Entries the caller may open",
"schema": {
"type": "object",
"additionalProperties": {
"type": "array",
"items": {
"$ref": "#/definitions/builtin.Entry"
}
}
}
},
"401": {
"description": "Authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/conversations/{conversationId}/selection/{addonId}": {
"get": {
"description": "Returns the resource the authenticated user selected for the conversation and add-on, with the user's current role and display name. The stored reference is keyed by the authenticated user, the conversation and the add-on, and is never a credential: the owner authorizes it again on every use. A reference that was revoked, deleted or never set is one answer: 404 not_selected. Another user's selection of the same conversation ID is never visible. Read-only. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Read the resource selected for a conversation",
"parameters": [
{
"type": "string",
"description": "Conversation ID",
"name": "conversationId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Add-on ID",
"name": "addonId",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "The selection",
"schema": {
"$ref": "#/definitions/gateway.Selection"
}
},
"400": {
"description": "Malformed conversation or add-on ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"404": {
"description": "Type not_selected",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"put": {
"description": "Stores which resource the authenticated user chose for the conversation and add-on, so that chat tool calls that name no resource use it. The caller must hold at least the viewer role on the resource now; the stored reference is never a bearer and is authorized again at the owner on every use. It is not process-global, not shared between users or conversations, and not part of the chat history. The store is bounded per user and expires unused entries. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Select the resource a conversation works on",
"parameters": [
{
"type": "string",
"description": "Conversation ID",
"name": "conversationId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Add-on ID",
"name": "addonId",
"in": "path",
"required": true
},
{
"description": "The resource",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.SelectionRequest"
}
}
],
"responses": {
"200": {
"description": "The selection",
"schema": {
"$ref": "#/definitions/gateway.Selection"
}
},
"400": {
"description": "Malformed IDs or body",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the resource is unknown or not readable by the caller",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"delete": {
"description": "Removes the authenticated user's selection for the conversation and add-on. Clearing a selection that does not exist succeeds. No MCP tool maps to this route.",
"tags": [
"Addons"
],
"summary": "Clear the resource selected for a conversation",
"parameters": [
{
"type": "string",
"description": "Conversation ID",
"name": "conversationId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Add-on ID",
"name": "addonId",
"in": "path",
"required": true
}
],
"responses": {
"204": {
"description": "Cleared"
},
"400": {
"description": "Malformed conversation or add-on ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/dispatch": {
"post": {
"description": "Authenticates the named caller, checks the addons feature, requires administration for every action except read, and forwards the typed request to the designated addon owner. A mutation is never retried: after an outcome_unknown response keep the operationId and poll operation-status instead of submitting a new mutation. Lifecycle actions are install, upgrade, enable, disable and uninstall (arguments sourceId and version; enable and disable take the generation that was listed). The management reads, all administrator-only and never mutations, are catalog-search (arguments text, kind, sourceId, offset, limit up to 100; kind is addon by default, or backend, or all), catalog-details (addonId plus arguments sourceId and version), catalog-sources and addons-installed. The library changes are catalog-source-save (arguments are one source: id, name, url, enabled, policy, authReference; removed true deletes it and never touches installed addons) and catalog-source-refresh (argument sourceId). A library names a credential only by a reference the owner configured, never a secret. Unknown argument fields are denied. The update actions are update-status and update-check (no arguments; the check fetches the library again and applies nothing) and update-upgrade (arguments expectedDigest and candidateDigest, the pinned upgrade); each answers with a payload {error: {type, message, status}} or {data}, the same from every frontend. The browser pages use the dedicated routes under /api/addons/{id}/update and /api/addons/{id}/upgrade, which accept only an administrator's own browser session.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Submit a typed addon lifecycle operation or read an addon resource",
"parameters": [
{
"description": "Typed addon request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/contract.Request"
}
}
],
"responses": {
"200": {
"description": "Operation accepted or read result",
"schema": {
"$ref": "#/definitions/contract.Result"
}
},
"400": {
"description": "Malformed or unknown request fields, or type invalid_request when the add-on refused the request as invalid: nothing was applied and the same call is refused again",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or owner policy denied the request, or type sdk_incompatible when an install is refused because the add-on needs another SDK version: the message names the range the add-on needs and the SDK version the host provides",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable, or the outcome of a mutation is unknown (type outcome_unknown, with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/pins": {
"get": {
"description": "Returns the ordered add-on IDs the caller pinned to the sidebar. The route never pins anything by itself; Studio and Talk are pinned once, at upgrade, for the accounts that existed then. The list holds only IDs the caller may see right now: a hidden or removed add-on is left out and the route does not reveal whether it exists. Needs an authenticated caller and no administrator role.",
"produces": [
"application/json"
],
"tags": [
"AddonNavigation"
],
"summary": "List the add-ons the caller pinned",
"responses": {
"200": {
"description": "Ordered add-on IDs",
"schema": {
"type": "array",
"items": {
"type": "string"
}
}
},
"401": {
"description": "Authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "The pin store is unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"put": {
"description": "Replaces the caller's pins with the ordered list in the body. IDs the caller may not see are dropped without any hint, and pins that are only hidden for now are kept. At most 32 IDs are accepted. Needs an authenticated caller from a same-origin browser request or a credential header, and no administrator role.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"AddonNavigation"
],
"summary": "Replace the add-ons the caller pinned",
"parameters": [
{
"description": "Ordered add-on IDs",
"name": "pins",
"in": "body",
"required": true,
"schema": {
"type": "array",
"items": {
"type": "string"
}
}
}
],
"responses": {
"200": {
"description": "The saved list the caller can see",
"schema": {
"type": "array",
"items": {
"type": "string"
}
}
},
"400": {
"description": "Not a list of strings, or longer than 32",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "The pin store or the add-on owner is unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals": {
"get": {
"description": "Returns the caller's proposals that still wait for a human decision, newest first, at most 32, each with its state, exact effect count, bounded sample, digest and expiry. This is where a change proposed by a chat tool, an agent or a scheduled job waits: only the proposer's own browser session can confirm it. A proposal of another user is never listed, and an expired one is left out. Read-only: it applies nothing. API keys may call it so an unattended agent can see what is waiting; they still cannot confirm.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "List the caller's own pending addon proposals",
"parameters": [
{
"enum": [
"pending"
],
"type": "string",
"description": "Only pending is supported",
"name": "state",
"in": "query",
"required": true
}
],
"responses": {
"200": {
"description": "The caller's pending proposals",
"schema": {
"$ref": "#/definitions/gateway.PendingList"
}
},
"400": {
"description": "Missing or unsupported state filter",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"post": {
"description": "A request with action create and no resourceId proposes the creation of a resource: the owner reserves a pending resource for the caller that nobody can reach, shows the requested name, and creates it only on confirmation by the proposer's browser session. Previews the mutation on the addon owner and stores it as a pending proposal with its exact effect count, a bounded sample, the revisions it saw and an expiry. Nothing is applied. The operationId becomes the proposal ID; repeating the same request returns the same proposal. Sharing and database deletion are never proposals. A named session or API key with the addons feature may propose; the proposal can only be applied by Confirm, which requires the proposer's own browser session. No MCP tool confirms a proposal.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Propose an addon data or schema mutation, or a resource creation, for confirmation",
"parameters": [
{
"description": "Typed write, schema or create request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/contract.Request"
}
}
],
"responses": {
"200": {
"description": "Stored proposal",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"400": {
"description": "Malformed or unknown request fields, or type invalid_request when the add-on refused the request as invalid: nothing was applied and the same call is refused again",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, resource policy or request binding denied the proposal. Type outcome_blocked (with blockedBy) when the caller is the owner role of the resource and another operation on it has an unresolved outcome; other callers get the plain denial",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable. Type outcome_unknown (with operationId): the proposal had been handed to the addon and its outcome is unknown, reconcile it. Type proposal_unavailable (with operationId): the owner could not serve this call and this call sent nothing to the addon; read the proposal for its state, which may already be settled, and repeat the same call with the same ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals/{id}": {
"get": {
"description": "Returns the caller's own proposal with its state, effects, receipt and, for a pending proposal that is blocked, the operation to resolve first. Expiry is recorded when a pending proposal is read after its expiry.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Read a stored addon proposal",
"parameters": [
{
"type": "string",
"description": "Proposal ID (the operation ID)",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Stored proposal",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the proposal is not the caller's",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals/{id}/cancel": {
"post": {
"description": "Withdraws the caller's pending proposal. A cancelled, expired or settled proposal can never run. Cancelling needs only a current identity.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Cancel a pending addon proposal",
"parameters": [
{
"type": "string",
"description": "Proposal ID (the operation ID)",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Cancelled proposal",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, the proposal is not the caller's, or it is not pending",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable. Type outcome_unknown (with operationId): the proposal had been handed to the addon and its outcome is unknown, reconcile it. Type proposal_unavailable (with operationId): the owner could not serve this call and this call sent nothing to the addon; read the proposal for its state, which may already be settled, and repeat the same call with the same ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals/{id}/confirm": {
"post": {
"description": "Applies exactly the stored operation once. This is the only way a stored mutation reaches an addon. It is accepted only from the proposer's own browser session: a same-origin fetch from the host UI that carries the X-LocalAI-Confirm header equal to the proposal ID and the digest of what was shown. API keys, bearer tokens, MCP sessions and requests that carry tool or agent execution markers are refused, and no MCP tool maps to this route. A repeated confirmation returns the stored receipt. If the reply is lost the outcome is outcome_unknown with the original operationId; resolve it with reconcile and never submit the change again.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Confirm and apply a pending addon proposal",
"parameters": [
{
"type": "string",
"description": "Proposal ID (the operation ID)",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Digest of the displayed proposal",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.ConfirmRequest"
}
}
],
"responses": {
"200": {
"description": "Proposal after the apply, with its receipt",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"400": {
"description": "Malformed or unknown request fields, or type invalid_request when the add-on refused the request as invalid: nothing was applied and the same call is refused again",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human confirmation surface, feature denied, or the proposal is not pending, is expired, stale, changed or blocked",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable. Type outcome_unknown (with operationId): the proposal had been handed to the addon and its outcome is unknown, reconcile it. Type proposal_unavailable (with operationId): the owner could not serve this call and this call sent nothing to the addon; read the proposal for its state, which may already be settled, and repeat the same call with the same ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals/{id}/reconcile": {
"post": {
"description": "Looks the operation's receipt up on the owner and records a committed or rejected result. It never applies anything and can be repeated. Use it after an outcome_unknown response. While an outcome is unresolved, other writes on the same resource are denied.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Resolve the outcome of an addon proposal by receipt lookup",
"parameters": [
{
"type": "string",
"description": "Proposal ID (the operation ID)",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Proposal after the lookup",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied or the proposal is not the caller's",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons, the owner or the addon runtime are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/proposals/{id}/resolve": {
"post": {
"description": "Explicit recovery for an operation whose receipt cannot be found. After outcome_unknown, repeated reconcile calls and the automatic lookups that did not settle it, the owner role of the resource may record that the outcome is not worth waiting for. The proposal becomes abandoned-unknown with the acting owner, the time and the stated reason, and stops blocking writes on the resource. The operation is never submitted again, and a later receipt lookup that says it ran is recorded as applied. The owner is asserting that waiting is not worth it: a change that did run stays in the data, and the revision check of later changes protects against applying on a stale view. Accepted only from a host confirmation surface with the same rules as confirm: the owner's own browser session, a same-origin fetch, the X-LocalAI-Confirm header equal to the proposal ID, and no tool, agent or MCP marker. Installation administrators and API keys have no access through this route, and no MCP tool maps to it.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Abandon an unresolved addon proposal outcome as the resource owner",
"parameters": [
{
"type": "string",
"description": "Proposal ID (the operation ID)",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Digest of the displayed proposal and the reason",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.ResolveRequest"
}
}
],
"responses": {
"200": {
"description": "Proposal after the acknowledgement, or settled by a final receipt",
"schema": {
"$ref": "#/definitions/gateway.Proposal"
}
},
"400": {
"description": "Malformed or unknown request fields, or type invalid_request when the add-on refused the request as invalid: nothing was applied and the same call is refused again",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human confirmation surface, not the resource owner, feature denied, digest changed or the proposal is not unresolved",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable. Always type proposal_unavailable (with operationId), whoever proposed the operation, because resolving is not an apply: the owner could not serve this call and it sent nothing to the addon; read the proposal for its state and repeat the same call with the same ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/relay": {
"post": {
"description": "Streams the response of the designated addon owner through the receiving frontend, bounded per kind by the configured byte and time limits. The kind selects the response: status (JSON), progress (text/event-stream), json, file (a private file result as an attachment: the handle is the artifact ID that a read returned, never an addon handle) or asset (text/javascript or text/css). An owner decision made before the first byte is returned with the same status from every frontend; a failure after the first byte aborts the stream.",
"consumes": [
"application/json"
],
"produces": [
"application/json",
"text/event-stream",
"application/octet-stream",
"text/javascript",
"text/css"
],
"tags": [
"Addons"
],
"summary": "Relay addon status, lifecycle progress, JSON, attachments or immutable UI assets",
"parameters": [
{
"description": "Relay request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.RelayRequest"
}
}
],
"responses": {
"200": {
"description": "Relayed response body",
"schema": {
"type": "string"
}
},
"400": {
"description": "Malformed or unknown request fields, or type invalid_request when the add-on refused the request as invalid: nothing was applied and the same call is refused again",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature or owner policy denied the request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources": {
"get": {
"description": "Returns the resources of one add-on that the caller owns or that were shared with the caller, each with the caller's own role, the display name the add-on reported (empty when it could not answer in time), the owner's display name for a shared resource and, for the owner role only, how many people it is shared with and the revision to send back with a rename or a delete. Resources of other people that are not shared with the caller are never listed, whatever the caller's installation role: installation administration is not data access. The answer also carries the installed generation, whether sharing is possible (false in local-owner mode) and the management operations the add-on supports. At most 100 resources are listed (truncated says more exist). Read-only. API keys may call it. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "List the add-on resources the caller may use",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "addonId",
"in": "query",
"required": true
}
],
"responses": {
"200": {
"description": "Resources the caller may use",
"schema": {
"$ref": "#/definitions/gateway.ResourceList"
}
},
"400": {
"description": "Missing or malformed add-on ID",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons, the owner or the add-on are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"post": {
"description": "Creates a resource through the add-on's create tool as the person's own explicit action, for example a new database. It needs no confirmation card: the request is the intent, and the same admission, quota, receipt and crash reconciliation rules as every mutation apply. The caller becomes the owner and the resource is private. Repeating the same operationId returns the same resource and never creates a second one; after outcome_unknown repeat the same call. The arguments are validated against the add-on's own tool definition. Accepted only from the person's own browser session: a same-origin fetch whose X-LocalAI-Confirm header equals the operationId, with no API key and no tool, agent or MCP marker. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Create an add-on resource (UI only)",
"parameters": [
{
"description": "Create request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.CreateResourceRequest"
}
}
],
"responses": {
"200": {
"description": "The created resource",
"schema": {
"$ref": "#/definitions/gateway.ResourceInfo"
}
},
"400": {
"description": "Malformed request or arguments the add-on's tool does not accept",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature or creation authority denied, or type quota_reached",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type stale: the generation is not the installed one",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId) after which the same call is repeated",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/apply": {
"post": {
"description": "Applies one change through a write or schema tool the add-on declares, for example inserting rows or adding a column, as the person's own explicit action in the person's own browser tab. It needs no confirmation card: the request is the intent. It is not a proposal and nothing is stored for later. The same rules as every change apply: the role of the action (the owner and editors write rows, only the owner changes the schema, viewers and installation administrators without a grant get the answer of an unknown resource), the declared tool definition, the resource's admission, the generation and the revision the caller read (409 stale, nothing changed), the add-on's receipt, which makes repeating the same operationId with the same request one change, and a wait behind an unresolved outcome. The tool's arguments must not name the resource, the tool, an actor or any field the tool definition does not declare. Reads, creations, the add-on's rename and delete, administrator tools and tools nobody declared are refused with 400. Accepted only from the person's own browser session: a same-origin fetch whose X-LocalAI-Confirm header equals the resource ID, with no API key and no tool, agent or MCP marker. After outcome_unknown, look the receipt up with kind=apply and the original operationId; never submit a new operation. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Change a resource's data or schema by the person's own action (UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"description": "Change request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.ApplyResourceRequest"
}
}
],
"responses": {
"200": {
"description": "The committed change and its receipt",
"schema": {
"$ref": "#/definitions/gateway.ApplyResult"
}
},
"400": {
"description": "Malformed request, a tool that is not a change, or arguments the tool does not accept",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, role too low or unknown resource, or type outcome_blocked",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type stale: the generation or the revision moved",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/delete": {
"post": {
"description": "Permanently deletes the resource, its data and every grant, as the owner's own explicit action. It is never a chat operation and never an uninstall. The request must carry the generation and revision the owner listed and confirm, which must equal the resource's display name exactly. The owner marks the resource as deleting (inaccessible to everyone, no new admission), stops and waits for admitted work, checks the revision, cancels pending proposals, asks the add-on to delete and only then removes the record and the grants. A crash leaves a deleting record that the owner finishes at startup; an unconfirmed one returns to use. A 503 with type outcome_unknown means the deletion is confirmed and being finished: repeat the same call, the resource is inaccessible meanwhile. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). No MCP tool maps to this route and the add-on's delete tool is not offered to any model.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Delete a resource and its data (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"description": "Delete request with the typed confirmation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.DeleteResourceRequest"
}
}
],
"responses": {
"200": {
"description": "Deleted",
"schema": {
"$ref": "#/definitions/gateway.DeleteResult"
}
},
"400": {
"description": "Malformed request, wrong confirmation or unsupported operation",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource, or type outcome_blocked",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type stale",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/grants": {
"get": {
"description": "Returns the grants of the whole resource: the immutable account ID, the display name and the role of each grantee. Only the owner role of the resource may read it; installation administrators, editors and viewers get the same denial as for an unknown resource. Local-owner mode has no sharing and answers 409 sharing_disabled. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "List who a resource is shared with (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Grants of the resource",
"schema": {
"$ref": "#/definitions/gateway.GrantList"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type sharing_disabled",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/grants/{userId}": {
"put": {
"description": "Gives the user the viewer or editor role on the whole resource; there are no per-table grants and the owner role cannot be granted. The user is named by the immutable account ID the user picker returned, and must be an active named account that may use add-ons. Lowering a role stops the user's admitted work and cancels the user's pending proposals before it is acknowledged. Local-owner mode answers 409 sharing_disabled. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). Safe to repeat; after outcome_unknown repeat the same call. No MCP tool maps to this route: sharing is never a tool.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Share a whole resource with a named user (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Account ID of the grantee",
"name": "userId",
"in": "path",
"required": true
},
{
"description": "Role",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.SetGrantRequest"
}
}
],
"responses": {
"200": {
"description": "Grants after the change",
"schema": {
"$ref": "#/definitions/gateway.GrantList"
}
},
"400": {
"description": "Malformed body or role",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role, unknown resource or a user that cannot be granted",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type sharing_disabled",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
},
"delete": {
"description": "Removes the user's grant. It returns only after the user's admitted work stopped and their pending proposals were cancelled; a drain that did not finish in the bound is a 503 and the revocation is already effective for new work. Local-owner mode answers 409 sharing_disabled. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). Safe to repeat. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Stop sharing a resource with a user (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Account ID of the grantee",
"name": "userId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Optional operation ID, returned with outcome_unknown",
"name": "operationId",
"in": "query"
}
],
"responses": {
"200": {
"description": "Grants after the change",
"schema": {
"$ref": "#/definitions/gateway.GrantList"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type sharing_disabled",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/operations/{operationId}": {
"get": {
"description": "Read-only receipt lookup after an outcome_unknown answer. kind is rename (operationId is the one the rename carried), delete or apply (the operationId of a change; any role that may write looks up its own). The state is pending, committed, rejected or unknown; it never applies anything. The owner role may look up a rename or a deletion and the roles that may write look up their own changes, from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). After a deletion finished the resource is gone and the lookup is denied like any unknown resource. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Look up the receipt of a rename or a deletion (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Operation ID of the rename",
"name": "operationId",
"in": "path",
"required": true
},
{
"enum": [
"rename",
"delete",
"apply"
],
"type": "string",
"description": "Operation kind",
"name": "kind",
"in": "query",
"required": true
},
{
"type": "string",
"description": "Apply only: the requestHash of the request the caller means (64 lowercase hex). A lookup whose hash is not the one the operation ID was used for is a 409 operation_conflict, never the other request's receipt",
"name": "requestHash",
"in": "query"
}
],
"responses": {
"200": {
"description": "Receipt state",
"schema": {
"$ref": "#/definitions/gateway.OperationStatus"
}
},
"400": {
"description": "Unsupported kind or malformed requestHash",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type operation_conflict: the operation ID belongs to another request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons, the owner or the add-on are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/proposals": {
"get": {
"description": "Returns the proposals on the resource whose outcome is unresolved (executing, outcome-unknown) or abandoned-unknown, whoever proposed them, so the resource owner can learn the operation ID and digest that resolve needs. Each entry has the proposal ID, state, the proposing user, the action, the exact effect count and the bounded effect sample the proposer saw, the expiry, the digest and the age in seconds. It never returns the proposed arguments or any credential. Read-only. Only the owner role of the resource may call it; installation administrators, editors and viewers are denied. It is accepted only from a host confirmation surface with the rules of resolve: the owner's own browser session, a same-origin fetch, the X-LocalAI-Confirm header equal to the resource ID, and no tool, agent or MCP marker. API keys are refused and no MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "List the unresolved addon proposals of a resource as its owner",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"enum": [
"unresolved"
],
"type": "string",
"description": "Only unresolved is supported",
"name": "state",
"in": "query",
"required": true
}
],
"responses": {
"200": {
"description": "Unresolved proposals, oldest first",
"schema": {
"$ref": "#/definitions/gateway.ProposalList"
}
},
"400": {
"description": "Missing or unsupported state filter",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human confirmation surface, not the owner role of the resource, feature denied or unknown resource",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/rename": {
"post": {
"description": "Renames the resource through the add-on as the owner's own explicit action, with no confirmation card. It still runs under the resource's admission, waits behind an unresolved outcome (403 outcome_blocked for the owner role), is checked against the generation and the revision the owner listed (409 stale, nothing changed) and leaves the add-on's receipt for the operation. Repeating the same operationId never renames twice; after outcome_unknown repeat the same call or look it up with the operation status route. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Rename a resource (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"description": "Rename request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/gateway.RenameResourceRequest"
}
}
],
"responses": {
"200": {
"description": "The renamed resource",
"schema": {
"$ref": "#/definitions/gateway.ResourceInfo"
}
},
"400": {
"description": "Malformed request or a name the add-on refuses",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource, or type outcome_blocked",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type stale",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or type outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/resources/{resourceId}/users": {
"get": {
"description": "Finds active named accounts that may use add-ons by a prefix of the display name or of the address, 2 to 64 characters, at most 20 results. Each result is an immutable account ID and a display name and nothing else. Only the owner role of the resource may search; the owner and the caller are left out. Local-owner mode answers 409 sharing_disabled. Accepted only from the owner's own browser session (same-origin fetch, X-LocalAI-Confirm equal to the resource ID). No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Search the people a resource can be shared with (owner, UI only)",
"parameters": [
{
"type": "string",
"description": "Resource ID",
"name": "resourceId",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Name or address prefix",
"name": "q",
"in": "query",
"required": true
},
{
"type": "integer",
"description": "At most this many results, up to 20",
"name": "limit",
"in": "query"
}
],
"responses": {
"200": {
"description": "Matching people",
"schema": {
"$ref": "#/definitions/gateway.UserList"
}
},
"400": {
"description": "Query too short or too long",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Not a human surface, feature denied, not the owner role or unknown resource",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type sharing_disabled",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/assets/{digest}/{path}": {
"get": {
"description": "Streams one file the addon manifest declares as UI code (the entry module or a style sheet) from the verified package on the owner, relayed through the receiving frontend. The digest in the path is the sha256 of the installed package, in lower-case hex, as listed in the ui entry of the apps list. The file is served only while the addon is installed and enabled and the digest is the current one; a file the manifest does not declare, a traversal, a directory, another addon, an older package, a disabled or uninstalled addon and a missing feature all give a denial without any file content, and a denial is never cached. The response is text/javascript or text/css with X-Content-Type-Options nosniff and Cache-Control private, max-age=31536000, immutable: the URL changes with the package, so the file never changes under it. The caller is authenticated by an API key header or by the same-origin browser session; a credential in the URL is never accepted. The addon process is never asked.",
"produces": [
"text/javascript",
"text/css"
],
"tags": [
"Addons"
],
"summary": "Read an immutable file of an addon UI package",
"parameters": [
{
"type": "string",
"description": "Addon ID",
"name": "id",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Package digest, 64 lower-case hex characters",
"name": "digest",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Declared package file, for example ui.js",
"name": "path",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "File bytes",
"schema": {
"type": "string"
},
"headers": {
"Cache-Control": {
"type": "string",
"description": "private, max-age=31536000, immutable"
},
"X-Content-Type-Options": {
"type": "string",
"description": "nosniff"
}
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the file, addon or package digest is not currently served",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/purge": {
"post": {
"description": "Administrators only, from their own browser session, with the add-on ID typed as confirm. Removes retained data the scope names: the add-on's data and the owner's records of its resources and shares (data), its backups (backups) and the data copies that restores and failed upgrades set aside (recoveryCopies). The add-on must be uninstalled or disabled. It is not the deletion of a database and not an uninstall; those never remove retained data. The purge is journaled and finished after a crash, and it touches no other add-on. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Purge the retained data of an add-on (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation, scope and confirmation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryPurgeRequest"
}
}
],
"responses": {
"200": {
"description": "What was purged, with warnings",
"schema": {
"$ref": "#/definitions/recovery.PurgeReceipt"
}
},
"400": {
"description": "Malformed request, empty scope or type confirmation_mismatch",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type addon_running, operation_conflict or busy",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery": {
"get": {
"description": "Administrators only. Returns the recovery state of the add-on's data (ready, upgrading, restoring, needs-recovery or purging) with its reason, the backups newest first with their size, creation time, reason (manual, upgrade or restore), the ownership mode and package version they hold and the last recorded verification result (verified, failed or unverified), the backup that is the recovery copy, the receipt and warnings of the last restore, the retention policy, the number of data copies a restore set aside and the operation IDs whose result is recorded. Read-only. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Read the data recovery state and the backups of an add-on (administrators)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Recovery state",
"schema": {
"$ref": "#/definitions/recovery.Status"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature or administration denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery/backup": {
"post": {
"description": "Administrators only, from their own browser session (same-origin fetch, X-LocalAI-Confirm equal to the operationId). Blocks new admission for the add-on's resources, invalidates its pending proposals, drains admitted work, stops the add-on, makes a consistent copy of its data (every SQLite file with VACUUM INTO, so the write-ahead log is part of it) together with the owner's access state, verifies the copy, publishes it atomically with restricted permissions and starts the add-on again if it was meant to run. A backup that cannot be completed leaves nothing behind. Repeating the operationId returns the same backup. 503 insufficient_storage means the free space did not fit the backup and nothing was changed. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Back up the data of an add-on (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryBackupRequest"
}
}
],
"responses": {
"200": {
"description": "The verified backup",
"schema": {
"$ref": "#/definitions/recovery.BackupReceipt"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"404": {
"description": "Type not_installed",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type busy",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, insufficient_storage, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery/delete-backup": {
"post": {
"description": "Administrators only, from their own browser session. The only verified copy and the recovery copy of an upgrade are never removed this way (409 sole_recovery_copy); a purge with explicit scope removes them. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Delete one backup of an add-on (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation and backup",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryDeleteBackupRequest"
}
}
],
"responses": {
"200": {
"description": "The deleted backup",
"schema": {
"$ref": "#/definitions/addons.RecoveryDeleted"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"404": {
"description": "Type backup_not_found",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type sole_recovery_copy or busy",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery/prune": {
"post": {
"description": "Administrators only, from their own browser session. Removes the backups the configured count and age policy no longer keeps. The newest verified backup and the recovery copy of an upgrade stay, whatever their age. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Apply the backup retention policy now (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryPruneRequest"
}
}
],
"responses": {
"200": {
"description": "The removed backups",
"schema": {
"$ref": "#/definitions/addons.RecoveryPruned"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type busy",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery/restore": {
"post": {
"description": "Administrators only, from their own browser session, with the add-on ID typed as confirm. The add-on must not be running. The backup is verified before anything changes, its files are staged and swapped in atomically, and the data they replace stays as a recovery copy until it is purged. Access is not restored as it was: every grant is checked again against the current accounts and the lowerings of access recorded since the backup, so nobody gets back access that was taken away; pending proposals are cancelled and operations with an unknown outcome are looked up by receipt and never applied again. The answer carries warnings: the time the data goes back to, the number of changes made after it that are lost, and the shares that were not restored. A backup made in local-owner mode needs mapping {\"local-owner\": \"\u003caccount id\u003e\"}; a backup of named-user data is refused in local-owner mode. Repeating the operationId returns the same receipt and does not restore again. No MCP tool maps to this route. The backup must belong to the add-on in the path; a backup of another add-on is refused with type invalid_request and nothing is touched.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Restore a backup of an add-on (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation, backup, mapping and confirmation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryRestoreRequest"
}
}
],
"responses": {
"200": {
"description": "What was restored, with warnings",
"schema": {
"$ref": "#/definitions/recovery.RestoreReceipt"
}
},
"400": {
"description": "Malformed request, type confirmation_mismatch or mapping_invalid",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"404": {
"description": "Type backup_not_found",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type addon_running, backup_corrupt, incompatible_format, identity_changed, mapping_required, downgrade_unsupported, operation_conflict or busy",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/recovery/verify": {
"post": {
"description": "Administrators only, from their own browser session. Reads every file of the backup against its manifest digests and runs the SQLite integrity check, and records the result shown in the list. A failed backup stays in place; removing it is a separate explicit action. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Verify a backup of an add-on (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation and backup",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.RecoveryVerifyRequest"
}
}
],
"responses": {
"200": {
"description": "The verified backup",
"schema": {
"$ref": "#/definitions/recovery.BackupReceipt"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"404": {
"description": "Type backup_not_found",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type backup_corrupt",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/update": {
"get": {
"description": "Returns the owner's update check for an installed add-on, for every authenticated caller who holds the addons feature and may see the add-on (an administrator any installed add-on, others only an enabled one; anything else is the same 403). The status is computed by the owner from the library the add-on was installed from and never from the browser: state is unknown, up-to-date, update-available, incompatible (with reasons: platform, host-version, api-version, protocol-version, sdk-version or data-format), check-failed (with a redacted error class and the time the library was last read) or integrity-error. A failed check is never up-to-date. A moving channel (development, a branch, latest) compares resolved digests, so a changed digest with the same version label is an update; an immutable version selects only a greater compatible version of the same major. The library and channel never change by themselves. The route never uses the network and never applies anything. Callers who are not administrators get the state, versions and times only, without library, channel or digests. No MCP tool maps to this route.",
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Read the update status of an addon",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Update status",
"schema": {
"$ref": "#/definitions/updates.Status"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature denied, or the add-on is unknown or hidden",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/update/check": {
"post": {
"description": "Administrators only, from their own browser session (same-origin fetch, X-LocalAI-Confirm equal to the operationId). The owner fetches the library the add-on was installed from again, through the credential reference the library is configured with, and answers with the new update status. It is rate limited per add-on and per library: within the interval the last status is returned with throttled true. It applies nothing; an upgrade is a separate explicit request. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Check an addon for updates now (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.UpdateCheckRequest"
}
}
],
"responses": {
"200": {
"description": "Update status",
"schema": {
"$ref": "#/definitions/updates.Status"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Addons or the owner are unavailable",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/addons/{id}/upgrade": {
"post": {
"description": "Administrators only, from their own browser session (same-origin fetch, X-LocalAI-Confirm equal to the operationId). The only action that applies an update. It carries the digest of the installed package the caller saw (expectedDigest) and the digest of the candidate the check resolved (candidateDigest). The owner re-resolves the library the add-on came from and refuses, before it records anything, an installed package that is not the expected one (409 stale_installed), a candidate that is not the one it resolves now (409 candidate_changed), an incompatible candidate (409 incompatible), a rewritten immutable version (409 integrity_error) and the same transition under another operation ID (409 upgrade_conflict). It never changes the library or the channel. The upgrade then follows the lifecycle of an add-on: the candidate is staged and verified, the data is backed up and migrated when the add-on has any, and the new generation is published only when it starts. A failure or cancellation keeps the last good generation or leaves an explicit recovery state; the previous installation is never deleted. Repeating the operationId returns the recorded result and never runs again; after outcome_unknown read operation-status with the same operationId. No MCP tool maps to this route.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"Addons"
],
"summary": "Upgrade an addon to the candidate that was checked (administrators, browser session)",
"parameters": [
{
"type": "string",
"description": "Add-on ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Operation and the pinned digests",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/addons.UpgradeRequest"
}
}
],
"responses": {
"200": {
"description": "Operation accepted; the payload is the operation status",
"schema": {
"$ref": "#/definitions/contract.Result"
}
},
"400": {
"description": "Malformed request",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"401": {
"description": "Named authentication required",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"403": {
"description": "Feature, administration or browser session denied",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"409": {
"description": "Type stale_installed, candidate_changed, incompatible, integrity_error, upgrade_conflict, upgrade_failed or operation_conflict",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
},
"503": {
"description": "Unavailable, check_failed, or outcome_unknown (with operationId)",
"schema": {
"$ref": "#/definitions/addons.AddonErrorResponse"
}
}
}
}
},
"/api/agent/jobs": {
"get": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "List agent jobs",
"parameters": [
{
"type": "string",
"description": "Filter by task ID",
"name": "task_id",
"in": "query"
},
{
"type": "string",
"description": "Filter by status (pending, running, completed, failed, cancelled)",
"name": "status",
"in": "query"
},
{
"type": "integer",
"description": "Max number of jobs to return",
"name": "limit",
"in": "query"
},
{
"type": "string",
"description": "Set to 'true' for admin cross-user listing",
"name": "all_users",
"in": "query"
}
],
"responses": {
"200": {
"description": "jobs",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Job"
}
}
}
}
}
},
"/api/agent/jobs/execute": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Execute an agent job",
"parameters": [
{
"description": "Job execution request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.JobExecutionRequest"
}
}
],
"responses": {
"201": {
"description": "job created",
"schema": {
"$ref": "#/definitions/schema.JobExecutionResponse"
}
},
"400": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/agent/jobs/{id}": {
"get": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Get an agent job",
"parameters": [
{
"type": "string",
"description": "Job ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "job",
"schema": {
"$ref": "#/definitions/schema.Job"
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Delete an agent job",
"parameters": [
{
"type": "string",
"description": "Job ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "message",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/agent/jobs/{id}/cancel": {
"post": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Cancel an agent job",
"parameters": [
{
"type": "string",
"description": "Job ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "message",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"400": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/agent/tasks": {
"get": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "List agent tasks",
"parameters": [
{
"type": "string",
"description": "Set to 'true' for admin cross-user listing",
"name": "all_users",
"in": "query"
}
],
"responses": {
"200": {
"description": "tasks",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Task"
}
}
}
}
},
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Create a new agent task",
"parameters": [
{
"description": "Task definition",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Task"
}
}
],
"responses": {
"201": {
"description": "id",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"400": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/agent/tasks/{id}": {
"get": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Get an agent task",
"parameters": [
{
"type": "string",
"description": "Task ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "task",
"schema": {
"$ref": "#/definitions/schema.Task"
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"put": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Update an agent task",
"parameters": [
{
"type": "string",
"description": "Task ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Updated task definition",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Task"
}
}
],
"responses": {
"200": {
"description": "message",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"400": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Delete an agent task",
"parameters": [
{
"type": "string",
"description": "Task ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "message",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/agent/tasks/{name}/execute": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"agent-jobs"
],
"summary": "Execute an agent task by name",
"parameters": [
{
"type": "string",
"description": "Task name",
"name": "name",
"in": "path",
"required": true
},
{
"description": "Optional template parameters",
"name": "parameters",
"in": "body",
"schema": {
"type": "object"
}
}
],
"responses": {
"201": {
"description": "job created",
"schema": {
"$ref": "#/definitions/schema.JobExecutionResponse"
}
},
"400": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/aliases": {
"get": {
"tags": [
"models"
],
"summary": "List model aliases",
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/localai.AliasInfo"
}
}
}
}
}
},
"/api/backend-logs": {
"get": {
"description": "Returns a sorted list of model IDs that have captured backend process output",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "List models with backend logs",
"responses": {
"200": {
"description": "Model IDs with logs",
"schema": {
"type": "array",
"items": {
"type": "string"
}
}
}
}
}
},
"/api/backend-logs/{modelId}": {
"get": {
"description": "Returns all captured log lines (stdout/stderr) for the specified model's backend process",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "Get backend logs for a model",
"parameters": [
{
"type": "string",
"description": "Model ID",
"name": "modelId",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Log lines",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/model.BackendLogLine"
}
}
}
}
}
},
"/api/backend-logs/{modelId}/clear": {
"post": {
"description": "Removes all captured log lines for the specified model's backend process",
"tags": [
"monitoring"
],
"summary": "Clear backend logs for a model",
"parameters": [
{
"type": "string",
"description": "Model ID",
"name": "modelId",
"in": "path",
"required": true
}
],
"responses": {
"204": {
"description": "Logs cleared"
}
}
}
},
"/api/backend-traces": {
"get": {
"description": "Returns a bounded, newest-first page of captured backend traces (LLM calls, embeddings, TTS, etc). The heavy body and data fields are omitted unless full=true; fetch them per-trace from /api/backend-traces/{id}. Paging metadata is returned in the X-Total-Count, X-Trace-Offset and X-Trace-Limit headers.",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "List backend operation traces",
"parameters": [
{
"type": "integer",
"description": "Maximum entries to return (default 50, max 1000, 0 for all)",
"name": "limit",
"in": "query"
},
{
"type": "integer",
"description": "Number of entries to skip (default 0)",
"name": "offset",
"in": "query"
},
{
"type": "boolean",
"description": "Include the body and data payloads (default false)",
"name": "full",
"in": "query"
}
],
"responses": {
"200": {
"description": "Backend operation traces",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/backend-traces/clear": {
"post": {
"description": "Removes all captured backend operation traces from the buffer",
"tags": [
"monitoring"
],
"summary": "Clear backend traces",
"responses": {
"204": {
"description": "Traces cleared"
}
}
}
},
"/api/backend-traces/{id}": {
"get": {
"description": "Returns a single captured backend trace, including the body and data payloads omitted from the list response",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "Get one backend operation trace",
"parameters": [
{
"type": "string",
"description": "Trace ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Backend operation trace",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "Trace not found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/branding": {
"get": {
"description": "Returns the configured instance name, tagline, and asset URLs. Public — no authentication required.",
"produces": [
"application/json"
],
"tags": [
"branding"
],
"summary": "Get instance branding",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.BrandingResponse"
}
}
}
}
},
"/api/branding/asset/{kind}": {
"post": {
"description": "Upload a custom logo, horizontal logo, or favicon. The file replaces any previous override for that kind.",
"consumes": [
"multipart/form-data"
],
"produces": [
"application/json"
],
"tags": [
"branding"
],
"summary": "Upload a branding asset",
"parameters": [
{
"type": "string",
"description": "Asset kind: logo, logo_horizontal, or favicon",
"name": "kind",
"in": "path",
"required": true
},
{
"type": "file",
"description": "Image file (png, jpeg, svg, webp, ico — up to 5MiB)",
"name": "file",
"in": "formData",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.BrandingResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"description": "Remove a custom branding asset; the UI falls back to the bundled LocalAI default.",
"produces": [
"application/json"
],
"tags": [
"branding"
],
"summary": "Reset a branding asset to default",
"parameters": [
{
"type": "string",
"description": "Asset kind: logo, logo_horizontal, or favicon",
"name": "kind",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.BrandingResponse"
}
}
}
}
},
"/api/conversations": {
"get": {
"description": "Returns the live conversations, newest first, and in ` + "`" + `deleted` + "`" + ` the conversations the user deleted. A client must drop any local copy whose id appears in ` + "`" + `deleted` + "`" + ` instead of uploading it again.",
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "List chat conversations",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.ConversationList"
}
}
}
},
"post": {
"description": "Stores the conversation and returns it with its new server-assigned ` + "`" + `version` + "`" + `. ` + "`" + `createdAt` + "`" + `/` + "`" + `updatedAt` + "`" + ` are kept as sent (Unix ms). Writing a conversation the user deleted returns 410 and does not bring it back.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Save a chat conversation (upsert)",
"parameters": [
{
"description": "Conversation payload",
"name": "body",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Conversation"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.Conversation"
}
},
"400": {
"description": "Invalid id or body",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"410": {
"description": "Deleted",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Delete all chat conversations for the current user",
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/conversations/bulk": {
"put": {
"description": "Upserts every conversation in the payload and deletes the user's other live conversations. Ids the user deleted earlier are not written and are listed in ` + "`" + `skipped` + "`" + `.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Replace all chat conversations",
"parameters": [
{
"description": "Conversations",
"name": "body",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ConversationBulkRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.ConversationBulkResponse"
}
},
"400": {
"description": "Invalid id or body",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/conversations/{id}": {
"get": {
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Get a chat conversation",
"parameters": [
{
"type": "string",
"description": "Conversation ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.Conversation"
}
},
"404": {
"description": "Never existed",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"410": {
"description": "Deleted",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"put": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Save a chat conversation by id (upsert)",
"parameters": [
{
"type": "string",
"description": "Conversation ID (overrides the body id)",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Conversation payload",
"name": "body",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.Conversation"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.Conversation"
}
},
"400": {
"description": "Invalid id or body",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"410": {
"description": "Deleted",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"description": "Deletes the conversation and keeps a tombstone so no client can write it back. Deleting an unknown or already deleted id succeeds.",
"produces": [
"application/json"
],
"tags": [
"chat-history"
],
"summary": "Delete a chat conversation",
"parameters": [
{
"type": "string",
"description": "Conversation ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"400": {
"description": "Invalid id",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/failover": {
"get": {
"produces": [
"application/json"
],
"tags": [
"failover"
],
"summary": "List failover chains and the health of their targets",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.FailoverChainsResponse"
}
}
}
}
},
"/api/failover/events": {
"get": {
"description": "The first event is \"snapshot\" with the full state, then \"chain.switched\" and \"target.state\" events.",
"produces": [
"text/event-stream"
],
"tags": [
"failover"
],
"summary": "Stream failover events (server-sent events)",
"responses": {
"200": {
"description": "OK"
}
}
}
},
"/api/failover/{chain}": {
"get": {
"produces": [
"application/json"
],
"tags": [
"failover"
],
"summary": "Get one failover chain",
"parameters": [
{
"type": "string",
"description": "Chain name",
"name": "chain",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/failover.ChainStatus"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/failover/{chain}/pin": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"failover"
],
"summary": "Pin a failover chain to one target",
"parameters": [
{
"type": "string",
"description": "Chain name",
"name": "chain",
"in": "path",
"required": true
},
{
"description": "Target to pin",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.FailoverPinRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/failover.ChainStatus"
}
},
"400": {
"description": "Bad Request",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
},
"delete": {
"produces": [
"application/json"
],
"tags": [
"failover"
],
"summary": "Remove the pin from a failover chain",
"parameters": [
{
"type": "string",
"description": "Chain name",
"name": "chain",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/failover.ChainStatus"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/fine-tuning/status": {
"get": {
"description": "Returns model_exports=false when this instance does not share the models path of the add-on owner. The value is the same as the model_exports key of GET /api/features.",
"produces": [
"application/json"
],
"tags": [
"fine-tuning",
"quantization"
],
"summary": "Report whether model exports are allowed",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.ModelExportsStatus"
}
},
"503": {
"description": "add-ons are off or the owner is not reachable",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/instructions": {
"get": {
"description": "Returns a compact list of instruction areas with descriptions and URLs for detailed guides",
"produces": [
"application/json"
],
"tags": [
"instructions"
],
"summary": "List available API instruction areas",
"responses": {
"200": {
"description": "instructions list with hint",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/instructions/{name}": {
"get": {
"description": "Returns a markdown guide (default) or filtered OpenAPI fragment (format=json) for a named instruction",
"produces": [
"application/json",
"text/markdown"
],
"tags": [
"instructions"
],
"summary": "Get an instruction's API guide or OpenAPI fragment",
"parameters": [
{
"type": "string",
"description": "Instruction name (e.g. chat-inference, config-management)",
"name": "name",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Response format: json for OpenAPI fragment, omit for markdown",
"name": "format",
"in": "query"
}
],
"responses": {
"200": {
"description": "instruction documentation",
"schema": {
"$ref": "#/definitions/localai.APIInstructionResponse"
}
},
"404": {
"description": "instruction not found",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/library/sources": {
"get": {
"description": "Every library source of the add-on, backend and model kinds, with its state: kind, id, name, url, enabled (what the host does with it), default (shipped), removed (a default the user removed; a tombstone), suspended (a default while the official libraries are off), the conflict message when it clashes with another source, and the last fetch state when there is one. Administrators only.",
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "List the library sources",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.LibrarySourcesResponse"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
},
"post": {
"description": "Adds a source of your own. kind is addon, backend or model; format is library (default) or gallery. The id is derived from the name or URL when it is empty. A source with the same kind and URL is refused (409), as is a taken id. Administrators only.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "Add a library source",
"responses": {
"201": {
"description": "Created",
"schema": {
"$ref": "#/definitions/library.Info"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/library/sources/{id}": {
"put": {
"description": "Replaces the mirrors of a model or backend source; an empty list clears them. Each mirror is an http(s), github:, oci:// or file:// address. Administrators only.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "Change the mirrors of a library source",
"parameters": [
{
"type": "string",
"description": "Source id",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/library.Info"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
},
"delete": {
"description": "A source of your own is deleted. A shipped default is not: it stays as a tombstone (removed true), so it does not come back at the next start; enabling it brings it back. Administrators only.",
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "Remove a library source",
"parameters": [
{
"type": "string",
"description": "Source id",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/library/sources/{id}/disable": {
"post": {
"description": "Enabling a removed default brings it back. Administrators only.",
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "Turn a library source on or off",
"parameters": [
{
"type": "string",
"description": "Source id",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/library.Info"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/library/sources/{id}/enable": {
"post": {
"description": "Enabling a removed default brings it back. Administrators only.",
"produces": [
"application/json"
],
"tags": [
"library"
],
"summary": "Turn a library source on or off",
"parameters": [
{
"type": "string",
"description": "Source id",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/library.Info"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/models/config-json/{name}": {
"patch": {
"description": "Deep-merges the JSON patch body into the existing model config",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"config"
],
"summary": "Partially update a model configuration",
"parameters": [
{
"type": "string",
"description": "Model name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "success message",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/models/config-metadata": {
"get": {
"description": "Returns config field metadata. Use ?section=\u003cid\u003e to filter by section, or omit for a section index.",
"produces": [
"application/json"
],
"tags": [
"config"
],
"summary": "List model configuration field metadata",
"parameters": [
{
"type": "string",
"description": "Section ID to filter (e.g. 'general', 'llm', 'parameters') or 'all' for everything",
"name": "section",
"in": "query"
}
],
"responses": {
"200": {
"description": "Section index or filtered field metadata",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/models/config-metadata/autocomplete/{provider}": {
"get": {
"description": "Returns runtime-resolved values for dynamic providers (backends, models)",
"produces": [
"application/json"
],
"tags": [
"config"
],
"summary": "Get dynamic autocomplete values for a config field",
"parameters": [
{
"type": "string",
"description": "Provider name (backends, models, models:chat, models:tts, models:transcript, models:vad)",
"name": "provider",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "values array",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/models/storage": {
"get": {
"description": "Per-model disk usage, shared-file relations and missing references under the models path",
"produces": [
"application/json"
],
"tags": [
"models"
],
"summary": "Report model storage usage on disk",
"responses": {
"200": {
"description": "storage report",
"schema": {
"$ref": "#/definitions/gallery.StorageIndex"
}
}
}
}
},
"/api/models/toggle-pinned/{name}/{action}": {
"put": {
"description": "Pin or unpin a model. Pinned models stay loaded and are excluded from automatic eviction.",
"tags": [
"config"
],
"summary": "Toggle model pinned status",
"parameters": [
{
"type": "string",
"description": "Model name",
"name": "name",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Action: 'pin' or 'unpin'",
"name": "action",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
}
}
}
},
"/api/models/vram-estimate": {
"post": {
"description": "Estimates VRAM based on model weight files at multiple context sizes",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"config"
],
"summary": "Estimate VRAM usage for a model",
"parameters": [
{
"description": "VRAM estimation parameters",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/modeladmin.VRAMRequest"
}
}
],
"responses": {
"200": {
"description": "VRAM estimate",
"schema": {
"$ref": "#/definitions/modeladmin.VRAMResponse"
}
}
}
}
},
"/api/models/{id}/load-cancel": {
"post": {
"description": "Cancels the load attempt named by ` + "`" + `job_id` + "`" + ` and stops its remote work. 200 means the attempt is gone or the worker confirmed the stop. 202 means the cancel is recorded and the stop is pending; the model is released after ` + "`" + `retry_after` + "`" + ` seconds regardless. 409 means a different attempt is current and carries its ` + "`" + `current_job_id` + "`" + `. Repeating the call is safe and does not extend the hold.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"models"
],
"summary": "Cancel one distributed model load.",
"parameters": [
{
"type": "string",
"description": "Model ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "The exact attempt to cancel",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ModelLoadCancelRequest"
}
}
],
"responses": {
"200": {
"description": "Stopped, or no such load any more",
"schema": {
"$ref": "#/definitions/schema.ModelLoadCancelResponse"
}
},
"202": {
"description": "Cancel recorded; stop pending",
"schema": {
"$ref": "#/definitions/schema.ModelLoadCancelResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"401": {
"description": "Unauthorized",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"403": {
"description": "Forbidden",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"404": {
"description": "Unknown model",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"409": {
"description": "A different attempt is current",
"schema": {
"$ref": "#/definitions/schema.ModelLoadCancelResponse"
}
}
}
}
},
"/api/models/{id}/load-status": {
"get": {
"description": "Returns the live state of a distributed cold load: job id, phase, node, byte progress, ETA, lease freshness, and for a failed attempt the cause, whether the stop is pending and when the model is released. 404 means no load exists for the model. A database error is 503, never an empty answer. This is the same ` + "`" + `loading` + "`" + ` object the 503 response carries while a model is still staging.",
"produces": [
"application/json"
],
"tags": [
"models"
],
"summary": "Report the progress of an in-flight model load.",
"parameters": [
{
"type": "string",
"description": "Model ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Live load progress",
"schema": {
"$ref": "#/definitions/schema.ModelLoadingStatus"
}
},
"404": {
"description": "No load is running for this model",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"503": {
"description": "The job table could not be read",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/models/{name}/{action}": {
"put": {
"description": "Enable or disable a model from being loaded on demand. Disabled models remain installed but cannot be loaded.",
"tags": [
"config"
],
"summary": "Toggle model enabled/disabled status",
"parameters": [
{
"type": "string",
"description": "Model name",
"name": "name",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Action: 'enable' or 'disable'",
"name": "action",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
}
}
}
},
"/api/motion/sessions": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"motion"
],
"summary": "Create a motion capture session",
"parameters": [
{
"description": "Model, output profile and source clock origin",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.MotionSessionRequest"
}
}
],
"responses": {
"201": {
"description": "Created",
"schema": {
"$ref": "#/definitions/localai.MotionSessionInfo"
}
}
}
}
},
"/api/motion/sessions/{id}": {
"get": {
"tags": [
"motion"
],
"summary": "Get motion session status",
"parameters": [
{
"type": "string",
"description": "Session ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.MotionSessionInfo"
}
}
}
},
"delete": {
"tags": [
"motion"
],
"summary": "Close a motion session",
"parameters": [
{
"type": "string",
"description": "Session ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"204": {
"description": "No Content"
}
}
}
},
"/api/motion/sessions/{id}/poses": {
"get": {
"tags": [
"motion"
],
"summary": "Stream motion frames and poses over a duplex WebSocket",
"parameters": [
{
"type": "string",
"description": "Session ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"101": {
"description": "Switching Protocols"
}
}
}
},
"/api/motion/sessions/{id}/tickets": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"motion"
],
"summary": "Issue a motion WebSocket ticket (30 seconds, single use)",
"parameters": [
{
"type": "string",
"description": "Session ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "Browser origin",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.MotionTicketRequest"
}
}
],
"responses": {
"201": {
"description": "Created",
"schema": {
"$ref": "#/definitions/auth.WebSocketTicketResponse"
}
}
}
}
},
"/api/nodes/models": {
"get": {
"tags": [
"Nodes"
],
"summary": "List all loaded models cluster-wide",
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/nodes.NodeModel"
}
}
}
}
}
},
"/api/nodes/{id}/max-replicas-per-model": {
"put": {
"tags": [
"Nodes"
],
"summary": "Update a node's max replicas per model",
"parameters": [
{
"type": "string",
"description": "Node ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "New value",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.UpdateMaxReplicasPerModelRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "integer"
}
}
},
"400": {
"description": "value must be \u003e= 1",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "node not found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
},
"delete": {
"tags": [
"Nodes"
],
"summary": "Reset a node's max replicas per model to the worker default",
"parameters": [
{
"type": "string",
"description": "Node ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "boolean"
}
}
},
"404": {
"description": "node not found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/nodes/{id}/vram-budget": {
"put": {
"tags": [
"Nodes"
],
"summary": "Update a node's VRAM allocation budget",
"parameters": [
{
"type": "string",
"description": "Node ID",
"name": "id",
"in": "path",
"required": true
},
{
"description": "New value (\\",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.UpdateVRAMBudgetRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"400": {
"description": "invalid budget",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "node not found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
},
"delete": {
"tags": [
"Nodes"
],
"summary": "Reset a node's VRAM budget to the worker default",
"parameters": [
{
"type": "string",
"description": "Node ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "object",
"additionalProperties": {
"type": "boolean"
}
}
},
"404": {
"description": "node not found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/p2p": {
"get": {
"tags": [
"p2p"
],
"summary": "Returns available P2P nodes",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.P2PNodesResponse"
}
}
}
}
}
},
"/api/p2p/token": {
"get": {
"tags": [
"p2p"
],
"summary": "Show the P2P token",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "string"
}
}
}
}
},
"/api/pii/analyze": {
"post": {
"description": "Runs the configured PII detectors (NER and/or pattern tiers) over the supplied text and returns the matched entity spans with the policy action that would fire. Detection only — the text is not modified and no block is enforced. Select detectors explicitly via ` + "`" + `detectors` + "`" + `, or pass a consuming ` + "`" + `model` + "`" + ` to use its effective policy: the model's own ` + "`" + `pii.detectors` + "`" + `, else the instance-wide ` + "`" + `pii_default_detectors` + "`" + `. A model with PII disabled, or enabled with nothing to scan with, is a 400. The raw matched value is never returned; admins may set ` + "`" + `reveal:true` + "`" + ` for the audit hash prefix.",
"tags": [
"pii"
],
"summary": "Detect PII entities in a string (no mutation).",
"parameters": [
{
"description": "text + detector selection",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.PIIAnalyzeRequest"
}
}
],
"responses": {
"200": {
"description": "Detected entities",
"schema": {
"$ref": "#/definitions/schema.PIIAnalyzeResponse"
}
}
}
}
},
"/api/pii/redact": {
"post": {
"description": "Runs the configured PII detectors over the text and applies each detector model's policy: masked spans are replaced with ` + "`" + `[REDACTED:\u003cid\u003e]` + "`" + `, allow spans pass through, and a single block action causes a 400 (type ` + "`" + `pii_blocked` + "`" + `) carrying the offending entities — the text is never returned in that case. Select detectors via ` + "`" + `detectors` + "`" + `, or a consuming ` + "`" + `model` + "`" + `'s effective policy (its own ` + "`" + `pii.detectors` + "`" + `, else the instance-wide ` + "`" + `pii_default_detectors` + "`" + `; PII must be enabled on the model). Records audit events (origin ` + "`" + `pii_redact` + "`" + `) visible at /api/pii/events.",
"tags": [
"pii"
],
"summary": "Redact PII in a string by applying the configured policy.",
"parameters": [
{
"description": "text + detector selection",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.PIIAnalyzeRequest"
}
}
],
"responses": {
"200": {
"description": "Redacted text + entities",
"schema": {
"$ref": "#/definitions/schema.PIIRedactResponse"
}
}
}
}
},
"/api/quantization/status": {
"get": {
"description": "Returns model_exports=false when this instance does not share the models path of the add-on owner. The value is the same as the model_exports key of GET /api/features.",
"produces": [
"application/json"
],
"tags": [
"fine-tuning",
"quantization"
],
"summary": "Report whether model imports are allowed",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.ModelExportsStatus"
}
},
"503": {
"description": "add-ons are off or the owner is not reachable",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/router/decide": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"router"
],
"summary": "Classify a prompt against a router model's policies (decision oracle)",
"parameters": [
{
"description": "decide params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.RouterDecideRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.RouterDecideResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"503": {
"description": "Service Unavailable",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/router/{name}/corpus": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"router"
],
"summary": "Seed the KNN routing corpus with labelled example prompts",
"parameters": [
{
"type": "string",
"description": "router model name",
"name": "name",
"in": "path",
"required": true
},
{
"description": "labelled exemplars",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.RouterCorpusAddRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.RouterCorpusAddResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
},
"delete": {
"produces": [
"application/json"
],
"tags": [
"router"
],
"summary": "Clear a router's KNN corpus",
"parameters": [
{
"type": "string",
"description": "router model name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.RouterCorpusClearResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/router/{name}/corpus/stats": {
"get": {
"produces": [
"application/json"
],
"tags": [
"router"
],
"summary": "Inspect a router's KNN corpus (label counts only, never texts)",
"parameters": [
{
"type": "string",
"description": "router model name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.RouterCorpusStatsResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/api/traces": {
"get": {
"description": "Returns a bounded, newest-first page of captured API exchange traces. Request and response bodies plus headers are omitted unless full=true; fetch them per-trace from /api/traces/{id}. Paging metadata is returned in the X-Total-Count, X-Trace-Offset and X-Trace-Limit headers.",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "List API request/response traces",
"parameters": [
{
"type": "integer",
"description": "Maximum entries to return (default 50, max 1000, 0 for all)",
"name": "limit",
"in": "query"
},
{
"type": "integer",
"description": "Number of entries to skip (default 0)",
"name": "offset",
"in": "query"
},
{
"type": "boolean",
"description": "Include request/response bodies and headers (default false)",
"name": "full",
"in": "query"
}
],
"responses": {
"200": {
"description": "Traced API exchanges",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/api/traces/clear": {
"post": {
"description": "Removes all captured API request/response traces from the buffer",
"tags": [
"monitoring"
],
"summary": "Clear API traces",
"responses": {
"204": {
"description": "Traces cleared"
}
}
}
},
"/api/traces/summary": {
"get": {
"description": "Returns request, failure and latency totals over a recent window, plus a bucketed series for sparklines. Exists so callers wanting three numbers do not have to fetch the whole trace list and count it themselves.",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "Summarize recent API traces",
"parameters": [
{
"type": "integer",
"description": "Window in hours (default 24, max 168)",
"name": "hours",
"in": "query"
}
],
"responses": {
"200": {
"description": "Counted trace totals",
"schema": {
"$ref": "#/definitions/middleware.TraceSummary"
}
}
}
}
},
"/api/traces/{id}": {
"get": {
"description": "Returns a single captured API exchange, including the request and response bodies omitted from the list response",
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "Get one API trace",
"parameters": [
{
"type": "string",
"description": "Trace ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Traced API exchange",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "Trace not found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/voice-profiles": {
"get": {
"description": "List saved voice-cloning references without exposing filesystem paths.",
"produces": [
"application/json"
],
"tags": [
"audio",
"voice-profiles"
],
"summary": "List voice profiles",
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.VoiceProfileListResponse"
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
},
"post": {
"description": "Save a consent-confirmed PCM WAV reference clip and exact transcript for voice cloning. Admin-only.",
"consumes": [
"multipart/form-data",
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"audio",
"voice-profiles"
],
"summary": "Create a voice profile",
"parameters": [
{
"type": "string",
"description": "Display name",
"name": "name",
"in": "formData",
"required": true
},
{
"type": "string",
"description": "Optional description",
"name": "description",
"in": "formData"
},
{
"type": "string",
"description": "Optional language tag",
"name": "language",
"in": "formData"
},
{
"type": "string",
"description": "Exact transcript of the reference clip",
"name": "transcript",
"in": "formData",
"required": true
},
{
"type": "boolean",
"description": "Confirms authorization to clone the voice",
"name": "consent_confirmed",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "16-bit PCM WAV, preferably mono 24 kHz, 1-120 seconds, up to 50 MiB",
"name": "audio",
"in": "formData",
"required": true
}
],
"responses": {
"201": {
"description": "Created",
"schema": {
"$ref": "#/definitions/voiceprofile.Profile"
}
},
"400": {
"description": "Bad Request",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"413": {
"description": "Request Entity Too Large",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/voice-profiles/{id}": {
"delete": {
"description": "Permanently remove a saved voice-cloning profile. Admin-only.",
"tags": [
"audio",
"voice-profiles"
],
"summary": "Delete a voice profile",
"parameters": [
{
"type": "string",
"description": "Voice profile UUID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"204": {
"description": "No Content"
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/api/voice-profiles/{id}/audio": {
"get": {
"description": "Stream the saved reference WAV for an authenticated TTS user.",
"produces": [
"audio/x-wav"
],
"tags": [
"audio",
"voice-profiles"
],
"summary": "Preview voice profile audio",
"parameters": [
{
"type": "string",
"description": "Voice profile UUID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"type": "string"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/audio/transform": {
"post": {
"description": "Runs an audio-in / audio-out transform conditioned on an optional auxiliary reference signal. Concrete transforms include AEC + noise suppression + dereverberation (LocalVQE), voice conversion (reference = target speaker), and pitch shifting. The backend determines the operation; pass model-specific tuning via repeated ` + "`" + `params[\u003ckey\u003e]=\u003cvalue\u003e` + "`" + ` form fields.",
"consumes": [
"multipart/form-data"
],
"produces": [
"audio/x-wav"
],
"tags": [
"audio"
],
"summary": "Transform audio (echo cancellation, noise suppression, voice conversion, etc.)",
"parameters": [
{
"type": "string",
"description": "model",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "primary input audio file",
"name": "audio",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "auxiliary reference audio (loopback for AEC, target voice for conversion, etc.)",
"name": "reference",
"in": "formData"
},
{
"type": "string",
"description": "wav | mp3 | ogg | flac",
"name": "response_format",
"in": "formData"
},
{
"type": "integer",
"description": "desired output sample rate in Hz; omit for the backend's own rate, otherwise 8000-192000",
"name": "sample_rate",
"in": "formData"
}
],
"responses": {
"200": {
"description": "transformed audio file",
"schema": {
"type": "string"
}
}
}
}
},
"/audio/transformations": {
"post": {
"description": "Runs an audio-in / audio-out transform conditioned on an optional auxiliary reference signal. Concrete transforms include AEC + noise suppression + dereverberation (LocalVQE), voice conversion (reference = target speaker), and pitch shifting. The backend determines the operation; pass model-specific tuning via repeated ` + "`" + `params[\u003ckey\u003e]=\u003cvalue\u003e` + "`" + ` form fields.",
"consumes": [
"multipart/form-data"
],
"produces": [
"audio/x-wav"
],
"tags": [
"audio"
],
"summary": "Transform audio (echo cancellation, noise suppression, voice conversion, etc.)",
"parameters": [
{
"type": "string",
"description": "model",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "primary input audio file",
"name": "audio",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "auxiliary reference audio (loopback for AEC, target voice for conversion, etc.)",
"name": "reference",
"in": "formData"
},
{
"type": "string",
"description": "wav | mp3 | ogg | flac",
"name": "response_format",
"in": "formData"
},
{
"type": "integer",
"description": "desired output sample rate in Hz; omit for the backend's own rate, otherwise 8000-192000",
"name": "sample_rate",
"in": "formData"
}
],
"responses": {
"200": {
"description": "transformed audio file",
"schema": {
"type": "string"
}
}
}
}
},
"/audio/transformations/stream": {
"get": {
"description": "Streams binary PCM frames in (interleaved stereo: ch0=audio, ch1=reference) and out (mono). The model must support the audio_transform use case. Any-to-any realtime_audio models use the OpenAI Realtime API instead. The first message must be a JSON ` + "`" + `session.update` + "`" + ` envelope describing model + sample format + frame size + backend params. Server emits binary PCM on the same cadence.",
"tags": [
"audio"
],
"summary": "Bidirectional realtime audio transform over WebSocket.",
"responses": {}
}
},
"/backend/load": {
"post": {
"description": "Loads the named model (or, for a realtime pipeline, all of its sub-models) into memory so subsequent requests pay no cold-start cost. The inverse of /backend/shutdown.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"monitoring"
],
"summary": "Pre-load a model into memory",
"parameters": [
{
"description": "Model to load",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ModelLoadRequest"
}
}
],
"responses": {
"200": {
"description": "Model loaded",
"schema": {
"$ref": "#/definitions/schema.ModelLoadResponse"
}
},
"400": {
"description": "Missing model name",
"schema": {
"$ref": "#/definitions/schema.ModelLoadResponse"
}
},
"500": {
"description": "Load failed (Loaded lists any sub-models that did load)",
"schema": {
"$ref": "#/definitions/schema.ModelLoadResponse"
}
}
}
}
},
"/backend/monitor": {
"get": {
"tags": [
"monitoring"
],
"summary": "Backend monitor endpoint",
"parameters": [
{
"type": "string",
"description": "Name of the model to monitor",
"name": "model",
"in": "query",
"required": true
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/proto.StatusResponse"
}
}
}
}
},
"/backend/shutdown": {
"post": {
"tags": [
"monitoring"
],
"summary": "Backend shutdown endpoint",
"parameters": [
{
"description": "Backend statistics request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.BackendMonitorRequest"
}
}
],
"responses": {}
}
},
"/backends": {
"get": {
"tags": [
"backends"
],
"summary": "List all Backends",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/gallery.GalleryBackend"
}
}
}
}
}
},
"/backends/apply": {
"post": {
"tags": [
"backends"
],
"summary": "Install backends to LocalAI.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.GalleryBackend"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.BackendResponse"
}
},
"403": {
"description": "Backend installation is disabled on this server (code backend_install_disabled)",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"409": {
"description": "Hardware-specific build in distributed mode (code concrete_backend_requires_target)",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/backends/available": {
"get": {
"tags": [
"backends"
],
"summary": "List all available Backends",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/gallery.GalleryBackend"
}
}
}
}
}
},
"/backends/delete/{name}": {
"post": {
"tags": [
"backends"
],
"summary": "delete backends from LocalAI.",
"parameters": [
{
"type": "string",
"description": "Backend name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.BackendResponse"
}
}
}
}
},
"/backends/galleries": {
"get": {
"tags": [
"backends"
],
"summary": "List all Galleries",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/config.Gallery"
}
}
}
}
}
},
"/backends/jobs": {
"get": {
"tags": [
"backends"
],
"summary": "Returns all the jobs status progress",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/galleryop.OpStatus"
}
}
}
}
}
},
"/backends/jobs/{uuid}": {
"get": {
"tags": [
"backends"
],
"summary": "Returns the job status",
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/galleryop.OpStatus"
}
}
}
}
},
"/backends/known": {
"get": {
"tags": [
"backends"
],
"summary": "List all known Backends (importer registry + curated pref-only + installed-on-disk)",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.KnownBackend"
}
}
}
}
}
},
"/backends/library/install": {
"post": {
"tags": [
"backends"
],
"summary": "Install a backend from a library entry on this node.",
"parameters": [
{
"description": "library, backend name and version",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.BackendLibraryInstallRequest"
}
}
],
"responses": {
"200": {
"description": "the job; read its status at the status URL",
"schema": {
"$ref": "#/definitions/schema.BackendResponse"
}
},
"400": {
"description": "the request does not name an entry",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"409": {
"description": "the same backend is already being installed",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/backends/library/updates": {
"get": {
"tags": [
"backends"
],
"summary": "List the installed library backends with a newer build in their library.",
"responses": {
"200": {
"description": "updates",
"schema": {
"type": "object",
"additionalProperties": {
"type": "array",
"items": {
"$ref": "#/definitions/backendlib.Update"
}
}
}
},
"503": {
"description": "on a frontend: the add-on owner is not reachable, does not serve library calls, or runs on another platform; the message says which",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/backends/upgrade/{name}": {
"post": {
"tags": [
"backends"
],
"summary": "Upgrade a backend",
"parameters": [
{
"type": "string",
"description": "Backend name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.BackendResponse"
}
},
"403": {
"description": "Backend installation is disabled on this server (code backend_install_disabled)",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/backends/upgrades": {
"get": {
"tags": [
"backends"
],
"summary": "Get available backend upgrades",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/gallery.UpgradeInfo"
}
}
}
}
}
},
"/backends/upgrades/check": {
"post": {
"tags": [
"backends"
],
"summary": "Force backend upgrade check",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/gallery.UpgradeInfo"
}
}
}
}
}
},
"/branding/asset/{kind}": {
"get": {
"description": "Serves the admin-uploaded logo, horizontal logo, or favicon. 404 when no override is set.",
"produces": [
"image/*"
],
"tags": [
"branding"
],
"summary": "Serve a custom branding asset",
"parameters": [
{
"type": "string",
"description": "Asset kind: logo, logo_horizontal, or favicon",
"name": "kind",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "OK"
},
"404": {
"description": "Not Found"
}
}
}
},
"/metrics": {
"get": {
"produces": [
"text/plain"
],
"tags": [
"monitoring"
],
"summary": "Prometheus metrics endpoint",
"responses": {
"200": {
"description": "Prometheus metrics",
"schema": {
"type": "string"
}
}
}
}
},
"/models/apply": {
"post": {
"tags": [
"models"
],
"summary": "Install models to LocalAI.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/localai.GalleryModel"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.GalleryResponse"
}
},
"409": {
"description": "The model's backend is not installed and backend installation is disabled (code backend_not_installed)",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/models/available": {
"get": {
"tags": [
"models"
],
"summary": "List installable models.",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/gallery.Metadata"
}
}
}
}
}
},
"/models/delete/{name}": {
"post": {
"tags": [
"models"
],
"summary": "delete models to LocalAI.",
"parameters": [
{
"type": "string",
"description": "Model name",
"name": "name",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.GalleryResponse"
}
}
}
}
},
"/models/galleries": {
"get": {
"tags": [
"models"
],
"summary": "List all Galleries",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "array",
"items": {
"$ref": "#/definitions/config.Gallery"
}
}
}
}
}
},
"/models/import": {
"post": {
"description": "Writes the posted model configuration. A config referencing\nremote assets (download_files entries, URI-valued model/mmproj)\nis enqueued on the gallery job queue and answered with a job id\nplus status URL to poll, like a gallery install; a config with\nno downloads is written synchronously and answered with a\nplain success response. A queued import runs the gallery\ninstaller, which also adds recommended inference defaults the\nconfig leaves unset and installs the backend when automatic\nbackend installation is enabled; the synchronous path stores\nthe body as posted.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"models"
],
"summary": "Import a model from a raw YAML or JSON configuration",
"parameters": [
{
"description": "model configuration (YAML or JSON body)",
"name": "request",
"in": "body",
"required": true,
"schema": {
"type": "object"
}
}
],
"responses": {
"200": {
"description": "remote assets present: job id and status URL",
"schema": {
"$ref": "#/definitions/schema.GalleryResponse"
}
},
"400": {
"description": "invalid or incomplete configuration",
"schema": {
"$ref": "#/definitions/localai.ModelResponse"
}
}
}
}
},
"/models/jobs": {
"get": {
"tags": [
"models"
],
"summary": "Returns all the jobs status progress",
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/galleryop.OpStatus"
}
}
}
}
}
},
"/models/jobs/{uuid}": {
"get": {
"tags": [
"models"
],
"summary": "Returns the job status",
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/galleryop.OpStatus"
}
}
}
}
},
"/system": {
"get": {
"tags": [
"monitoring"
],
"summary": "Show the LocalAI instance information",
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.SystemInformationResponse"
}
}
}
}
},
"/tokenMetrics": {
"get": {
"consumes": [
"application/json"
],
"produces": [
"audio/x-wav"
],
"tags": [
"tokenize"
],
"summary": "Get TokenMetrics for Active Slot.",
"responses": {
"200": {
"description": "generated audio/wav file",
"schema": {
"type": "string"
}
}
}
}
},
"/tts": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"audio/x-wav"
],
"tags": [
"audio"
],
"summary": "Generates audio from the input text.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.TTSRequest"
}
}
],
"responses": {
"200": {
"description": "generated audio/wav file",
"schema": {
"type": "string"
}
}
}
}
},
"/v1/audio/classification": {
"post": {
"consumes": [
"multipart/form-data"
],
"tags": [
"audio"
],
"summary": "Classify sound events in audio (audio tagging).",
"parameters": [
{
"type": "string",
"description": "model",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "audio file",
"name": "file",
"in": "formData",
"required": true
},
{
"type": "integer",
"description": "number of top tags to return (0 = backend default)",
"name": "top_k",
"in": "formData"
},
{
"type": "number",
"description": "drop tags scoring below this value",
"name": "threshold",
"in": "formData"
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.SoundClassificationResult"
}
}
}
}
},
"/v1/audio/diarization": {
"post": {
"description": "JSON accepts model, file (raw base64 audio), include_text, include_speaker_profiles, include_sounds and response_format. Profiles require voice-recognition permission and json or verbose_json; unsupported backends return 501.",
"consumes": [
"multipart/form-data",
"application/json"
],
"tags": [
"audio"
],
"summary": "Identify speakers in audio (who spoke when).",
"parameters": [
{
"type": "string",
"description": "model",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "audio file",
"name": "file",
"in": "formData",
"required": true
},
{
"type": "integer",
"description": "exact speaker count (\u003e0 forces; 0 = auto)",
"name": "num_speakers",
"in": "formData"
},
{
"type": "integer",
"description": "lower bound when auto-detecting",
"name": "min_speakers",
"in": "formData"
},
{
"type": "integer",
"description": "upper bound when auto-detecting",
"name": "max_speakers",
"in": "formData"
},
{
"type": "number",
"description": "clustering distance threshold when num_speakers is unknown",
"name": "clustering_threshold",
"in": "formData"
},
{
"type": "number",
"description": "discard segments shorter than this (seconds)",
"name": "min_duration_on",
"in": "formData"
},
{
"type": "number",
"description": "merge gaps shorter than this (seconds)",
"name": "min_duration_off",
"in": "formData"
},
{
"type": "string",
"description": "audio language hint (only meaningful for backends that bundle ASR)",
"name": "language",
"in": "formData"
},
{
"type": "boolean",
"description": "export portable biometric profiles (voice-recognition permission; JSON formats only)",
"name": "include_speaker_profiles",
"in": "formData"
},
{
"type": "boolean",
"description": "include per-segment transcript when the backend supports it",
"name": "include_text",
"in": "formData"
},
{
"type": "boolean",
"description": "include closed sound events (start, end, label, confidence) when the model has a sound_model companion; otherwise 501 include_sounds_unsupported (JSON formats only)",
"name": "include_sounds",
"in": "formData"
},
{
"type": "string",
"description": "json (default), verbose_json, or rttm",
"name": "response_format",
"in": "formData"
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.DiarizationResult"
}
}
}
}
},
"/v1/audio/speech": {
"post": {
"consumes": [
"application/json"
],
"produces": [
"audio/x-wav"
],
"tags": [
"audio"
],
"summary": "Generates audio from the input text.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.TTSRequest"
}
}
],
"responses": {
"200": {
"description": "generated audio/wav file",
"schema": {
"type": "string"
}
}
}
}
},
"/v1/audio/transcriptions": {
"post": {
"consumes": [
"multipart/form-data"
],
"tags": [
"audio"
],
"summary": "Transcribes audio into the input language.",
"parameters": [
{
"type": "string",
"description": "model",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "file",
"name": "file",
"in": "formData",
"required": true
},
{
"type": "number",
"description": "sampling temperature",
"name": "temperature",
"in": "formData"
},
{
"type": "array",
"items": {
"type": "string"
},
"collectionFormat": "csv",
"description": "timestamp granularities (word, segment)",
"name": "timestamp_granularities",
"in": "formData"
},
{
"type": "boolean",
"description": "stream partial results as SSE",
"name": "stream",
"in": "formData"
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/v1/audio/voices": {
"get": {
"description": "List named voices and their language and gender metadata. Use the optional model query parameter to filter the response.",
"produces": [
"application/json"
],
"tags": [
"audio"
],
"summary": "List text-to-speech voices",
"parameters": [
{
"type": "string",
"description": "Installed model name",
"name": "model",
"in": "query"
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/localai.TTSVoicesResponse"
}
},
"404": {
"description": "Not Found",
"schema": {
"$ref": "#/definitions/schema.ErrorResponse"
}
}
}
}
},
"/v1/chat/completions": {
"post": {
"tags": [
"inference"
],
"summary": "Generate a chat completions for a given prompt and model.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/completions": {
"post": {
"tags": [
"inference"
],
"summary": "Generate completions for a given prompt and model.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/decisions": {
"post": {
"description": "Uses the same local decision and NER execution as SystemOne. Supports inline images on capable backends; confidence is not OpenAI calibrated.",
"consumes": [
"application/json"
],
"produces": [
"application/json"
],
"tags": [
"systemone"
],
"summary": "Answer ordered predicate, choice and score questions.",
"parameters": [
{
"description": "input + questions + model",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.DecisionsRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.DecisionsResponse"
}
}
}
}
},
"/v1/depth": {
"post": {
"tags": [
"depth"
],
"summary": "Estimates per-pixel depth (and optionally pose/points) from an image.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.DepthRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.DepthResponse"
}
}
}
}
},
"/v1/detection": {
"post": {
"tags": [
"detection"
],
"summary": "Detects objects in the input image.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.DetectionRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.DetectionResponse"
}
}
}
}
},
"/v1/detokenize": {
"post": {
"tags": [
"tokenize"
],
"summary": "Detokenize the input.",
"parameters": [
{
"description": "Request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.DetokenizeRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.DetokenizeResponse"
}
}
}
}
},
"/v1/edits": {
"post": {
"tags": [
"inference"
],
"summary": "OpenAI edit endpoint",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/embeddings": {
"post": {
"tags": [
"embeddings"
],
"summary": "Get a vector representation of a given input that can be easily consumed by machine learning models and algorithms.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/face/analyze": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Analyze demographic attributes (age, gender, ...) of faces.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceAnalyzeRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.FaceAnalyzeResponse"
}
}
}
}
},
"/v1/face/embed": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Extract a face embedding from an image.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceEmbedRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.FaceEmbedResponse"
}
}
}
}
},
"/v1/face/forget": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Remove a previously-registered face by ID.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceForgetRequest"
}
}
],
"responses": {
"204": {
"description": "No Content"
}
}
}
},
"/v1/face/identify": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Identify a face against the registered database (1:N recognition).",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceIdentifyRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.FaceIdentifyResponse"
}
}
}
}
},
"/v1/face/register": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Register a face for 1:N identification.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceRegisterRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.FaceRegisterResponse"
}
}
}
}
},
"/v1/face/verify": {
"post": {
"tags": [
"face-recognition"
],
"summary": "Verify that two images depict the same person.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.FaceVerifyRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.FaceVerifyResponse"
}
}
}
}
},
"/v1/forecast": {
"post": {
"tags": [
"forecast"
],
"summary": "Forecasts numeric time series as quantile trajectories.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ForecastRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ForecastResponse"
}
}
}
}
},
"/v1/images/generations": {
"post": {
"tags": [
"images"
],
"summary": "Creates an image given a prompt.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/images/inpainting": {
"post": {
"description": "Perform image inpainting. Accepts multipart/form-data with ` + "`" + `image` + "`" + ` and ` + "`" + `mask` + "`" + ` files.",
"consumes": [
"multipart/form-data"
],
"produces": [
"application/json"
],
"tags": [
"images"
],
"summary": "Image inpainting",
"parameters": [
{
"type": "string",
"description": "Model identifier",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "string",
"description": "Text prompt guiding the generation",
"name": "prompt",
"in": "formData",
"required": true
},
{
"type": "integer",
"description": "Number of inference steps (default 25)",
"name": "steps",
"in": "formData"
},
{
"type": "file",
"description": "Original image file",
"name": "image",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "Mask image file (white = area to inpaint)",
"name": "mask",
"in": "formData",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/v1/images/upscale": {
"post": {
"description": "Upscale an image using a specified model (e.g. stable-diffusion-x4-upscaler). Accepts multipart/form-data.",
"consumes": [
"multipart/form-data"
],
"produces": [
"application/json"
],
"tags": [
"images"
],
"summary": "Image upscaling",
"parameters": [
{
"type": "string",
"description": "Upscaler model identifier (e.g. stable-diffusion-x4-upscaler)",
"name": "model",
"in": "formData",
"required": true
},
{
"type": "file",
"description": "Input image file",
"name": "image",
"in": "formData",
"required": true
},
{
"type": "integer",
"description": "Upscale factor: integer from 1 to 16; must match the model native scale",
"name": "scale",
"in": "formData",
"required": true
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
},
"500": {
"description": "Internal Server Error",
"schema": {
"type": "object",
"additionalProperties": {
"type": "string"
}
}
}
}
}
},
"/v1/mcp/chat/completions": {
"post": {
"tags": [
"mcp"
],
"summary": "MCP chat completions with automatic tool execution",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenAIRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/v1/messages": {
"post": {
"tags": [
"inference"
],
"summary": "Generate a message response for the given messages and model.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.AnthropicRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.AnthropicResponse"
}
}
}
}
},
"/v1/models": {
"get": {
"tags": [
"models"
],
"summary": "List and describe the various models available in the API.",
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ModelsDataResponse"
}
}
}
}
},
"/v1/models/capabilities": {
"get": {
"tags": [
"models"
],
"summary": "List available models enriched with capabilities and input/output modalities.",
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ModelCapabilitiesResponse"
}
}
}
}
},
"/v1/moderations": {
"post": {
"tags": [
"moderation"
],
"summary": "Classify text for potentially harmful content.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ModerationRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ModerationResponse"
}
}
}
}
},
"/v1/rerank": {
"post": {
"tags": [
"rerank"
],
"summary": "Reranks a list of phrases by relevance to a given text query.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.JINARerankRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.JINARerankResponse"
}
}
}
}
},
"/v1/responses": {
"post": {
"tags": [
"inference"
],
"summary": "Create a response using the Open Responses API",
"parameters": [
{
"description": "Request body",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.OpenResponsesRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ORResponseResource"
}
}
}
}
},
"/v1/responses/{id}": {
"get": {
"description": "Retrieve a response by ID. Can be used for polling background responses or resuming streaming responses.",
"tags": [
"inference"
],
"summary": "Get a response by ID",
"parameters": [
{
"type": "string",
"description": "Response ID",
"name": "id",
"in": "path",
"required": true
},
{
"type": "string",
"description": "Set to 'true' to resume streaming",
"name": "stream",
"in": "query"
},
{
"type": "integer",
"description": "Sequence number to resume from (for streaming)",
"name": "starting_after",
"in": "query"
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ORResponseResource"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/v1/responses/{id}/cancel": {
"post": {
"description": "Cancel a background response if it's still in progress",
"tags": [
"inference"
],
"summary": "Cancel a response",
"parameters": [
{
"type": "string",
"description": "Response ID",
"name": "id",
"in": "path",
"required": true
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.ORResponseResource"
}
},
"400": {
"description": "Bad Request",
"schema": {
"type": "object",
"additionalProperties": true
}
},
"404": {
"description": "Not Found",
"schema": {
"type": "object",
"additionalProperties": true
}
}
}
}
},
"/v1/slots/restore": {
"post": {
"description": "Loads the prompt cache that a save call returned an ID for into one slot of a llama-cpp model. The ID must have been issued for this model. The state is only as fresh as the save. Admin only.",
"tags": [
"inference"
],
"summary": "Restore a slot's prompt cache by ID.",
"parameters": [
{
"description": "Request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.SlotCacheRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.SlotCacheResponse"
}
}
}
}
},
"/v1/slots/save": {
"post": {
"description": "Writes the prompt cache (KV state) of one slot of a llama-cpp model under the directory set by the model's slot_save_path option, and returns an ID for it. The server issues the ID; the client cannot choose a name or a path. The server keeps at most 32 saves per model and deletes the oldest one beyond that. Admin only.",
"tags": [
"inference"
],
"summary": "Save a slot's prompt cache.",
"parameters": [
{
"description": "Request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.SlotCacheRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.SlotCacheResponse"
}
}
}
}
},
"/v1/sound-generation": {
"post": {
"tags": [
"audio"
],
"summary": "Generates audio from the input text.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.ElevenLabsSoundGenerationRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "string"
}
}
}
}
},
"/v1/systemone": {
"post": {
"description": "Runs zero-shot NER over the supplied state and answers each question. Question types: noul (binary entity presence), choice (pick one option), score (pick one level).",
"tags": [
"systemone"
],
"summary": "Answer structured-extraction questions over state text.",
"parameters": [
{
"description": "state + questions",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.SystemOneRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.SystemOneResponse"
}
}
}
}
},
"/v1/systemone/permute": {
"post": {
"description": "Re-runs one choice question under n_perm option orders. Reports per-order probabilities, argmax stability, and spread.",
"tags": [
"systemone"
],
"summary": "Re-run a choice question under multiple option orders.",
"parameters": [
{
"description": "request + question + n_perm + seed",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.SystemOnePermuteRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.SystemOnePermuteResponse"
}
}
}
}
},
"/v1/systemone/separate": {
"post": {
"description": "Runs N independent NER passes, one per question, against the same state. Response shape matches /v1/systemone.",
"tags": [
"systemone"
],
"summary": "Answer each question in a separate NER pass.",
"parameters": [
{
"description": "state + questions",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.SystemOneRequest"
}
}
],
"responses": {
"200": {
"description": "OK",
"schema": {
"$ref": "#/definitions/schema.SystemOneResponse"
}
}
}
}
},
"/v1/text-to-speech/{voice-id}": {
"post": {
"tags": [
"audio"
],
"summary": "Generates audio from the input text.",
"parameters": [
{
"type": "string",
"description": "Account ID",
"name": "voice-id",
"in": "path",
"required": true
},
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.TTSRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"type": "string"
}
}
}
}
},
"/v1/tokenMetrics": {
"get": {
"consumes": [
"application/json"
],
"produces": [
"audio/x-wav"
],
"tags": [
"tokenize"
],
"summary": "Get TokenMetrics for Active Slot.",
"responses": {
"200": {
"description": "generated audio/wav file",
"schema": {
"type": "string"
}
}
}
}
},
"/v1/tokenize": {
"post": {
"tags": [
"tokenize"
],
"summary": "Tokenize the input.",
"parameters": [
{
"description": "Request",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.TokenizeRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.TokenizeResponse"
}
}
}
}
},
"/v1/voice/analyze": {
"post": {
"tags": [
"voice-recognition"
],
"summary": "Analyze demographic attributes (age, gender, emotion) from a voice clip.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceAnalyzeRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.VoiceAnalyzeResponse"
}
}
}
}
},
"/v1/voice/embed": {
"post": {
"tags": [
"voice-recognition"
],
"summary": "Extract a speaker embedding from an audio clip.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceEmbedRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.VoiceEmbedResponse"
}
}
}
}
},
"/v1/voice/forget": {
"post": {
"tags": [
"voice-recognition"
],
"summary": "Remove a previously-registered speaker by ID.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceForgetRequest"
}
}
],
"responses": {
"204": {
"description": "No Content"
}
}
}
},
"/v1/voice/identify": {
"post": {
"tags": [
"voice-recognition"
],
"summary": "Identify a speaker against the registered database (1:N recognition).",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceIdentifyRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.VoiceIdentifyResponse"
}
}
}
}
},
"/v1/voice/register": {
"post": {
"description": "Supply either audio or speaker_profiles plus an explicit numeric speaker_slot. The selected model must expose matching trusted encoder metadata for portable enrollment. Registrations are global and ephemeral, with a fresh ID for each request.",
"tags": [
"voice-recognition"
],
"summary": "Register a speaker for 1:N identification.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceRegisterRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.VoiceRegisterResponse"
}
}
}
}
},
"/v1/voice/verify": {
"post": {
"tags": [
"voice-recognition"
],
"summary": "Verify that two audio clips were spoken by the same person.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VoiceVerifyRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.VoiceVerifyResponse"
}
}
}
}
},
"/vad": {
"post": {
"consumes": [
"application/json"
],
"tags": [
"audio"
],
"summary": "Detect voice fragments in an audio stream",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VADRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/proto.VADResponse"
}
}
}
}
},
"/video": {
"post": {
"tags": [
"video"
],
"summary": "Creates a video from a prompt and optional image or audio conditioning.",
"parameters": [
{
"description": "query params",
"name": "request",
"in": "body",
"required": true,
"schema": {
"$ref": "#/definitions/schema.VideoRequest"
}
}
],
"responses": {
"200": {
"description": "Response",
"schema": {
"$ref": "#/definitions/schema.OpenAIResponse"
}
}
}
}
},
"/ws/backend-logs/{modelId}": {
"get": {
"description": "Opens a WebSocket connection for real-time backend log streaming. Sends an initial batch of existing lines (type \"initial\"), then streams new lines as they appear (type \"line\"). Supports ping/pong keepalive.",
"tags": [
"monitoring"
],
"summary": "Stream backend logs via WebSocket",
"parameters": [
{
"type": "string",
"description": "Model ID",
"name": "modelId",
"in": "path",
"required": true
}
],
"responses": {}
}
}
},
"definitions": {
"addons.AddonError": {
"type": "object",
"properties": {
"blockedBy": {
"description": "BlockedBy is the unresolved operation that blocks the write. It is\npresent only with type outcome_blocked, and only for the owner role of\nthe resource.",
"type": "string"
},
"code": {
"description": "Code repeats the HTTP status. It is absent from outcome_unknown responses.",
"type": "integer"
},
"message": {
"description": "Message describes the failure and never carries owner details.",
"type": "string"
},
"operationId": {
"description": "OperationID is the original operation, which is the proposal ID for a\nproposal call. It is present only with type outcome_unknown or\nproposal_unavailable.",
"type": "string"
},
"type": {
"description": "Type is addon_error, outcome_unknown when a mutation may have run,\nproposal_unavailable when a proposal call could not be served and the\nsame call can be repeated by the same proposal ID after reading it, or\noutcome_blocked (403) when another operation on the resource has an\nunresolved outcome and the caller is the resource's owner role. The\nresource management routes add sharing_disabled (409, local-owner mode\nhas no named users to share with), stale (409, the generation or the\nrevision the caller saw is out of date), operation_conflict (409, the\noperation ID was used for another request and nothing was applied),\ninvalid_request (400, for\nexample a missing deletion confirmation or an unsupported operation),\nquota_reached (403) and not_selected (404, no resource is selected for\nthe conversation).",
"type": "string",
"example": "outcome_unknown"
}
}
},
"addons.AddonErrorResponse": {
"type": "object",
"properties": {
"error": {
"$ref": "#/definitions/addons.AddonError"
}
}
},
"addons.ConfirmRequest": {
"type": "object",
"properties": {
"digest": {
"description": "Digest is the digest of the proposal as it was displayed.",
"type": "string"
}
}
},
"addons.RecoveryBackupRequest": {
"type": "object",
"properties": {
"operationId": {
"description": "OperationID identifies the request. Repeating it returns the backup it\nmade and makes no second one.",
"type": "string"
}
}
},
"addons.RecoveryDeleteBackupRequest": {
"type": "object",
"properties": {
"backupId": {
"type": "string"
},
"operationId": {
"type": "string"
}
}
},
"addons.RecoveryDeleted": {
"type": "object",
"properties": {
"deleted": {
"type": "string"
}
}
},
"addons.RecoveryPruneRequest": {
"type": "object",
"properties": {
"operationId": {
"type": "string"
}
}
},
"addons.RecoveryPruned": {
"type": "object",
"properties": {
"removed": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"addons.RecoveryPurgeRequest": {
"type": "object",
"properties": {
"confirm": {
"description": "Confirm is the add-on's ID, typed.",
"type": "string"
},
"operationId": {
"type": "string"
},
"scope": {
"$ref": "#/definitions/recovery.PurgeScope"
}
}
},
"addons.RecoveryRestoreRequest": {
"type": "object",
"properties": {
"backupId": {
"type": "string"
},
"confirm": {
"description": "Confirm is the add-on's ID, typed.",
"type": "string"
},
"mapping": {
"description": "Mapping names the current account that takes the data of a backup made\nin another ownership mode, as {\"local-owner\": \"\u003caccount id\u003e\"}. It is\nexplicit: there is no wildcard and no default.",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"operationId": {
"type": "string"
}
}
},
"addons.RecoveryVerifyRequest": {
"type": "object",
"properties": {
"backupId": {
"type": "string"
},
"operationId": {
"type": "string"
}
}
},
"addons.ResolveRequest": {
"type": "object",
"properties": {
"digest": {
"description": "Digest is the digest of the proposal as it was displayed.",
"type": "string"
},
"reason": {
"description": "Reason is why the owner stops waiting. Required, at most 512 characters.",
"type": "string"
}
}
},
"addons.UpdateCheckRequest": {
"type": "object",
"properties": {
"operationId": {
"description": "OperationID identifies the request and is the confirmation header.",
"type": "string",
"example": "check-notes-1"
}
}
},
"addons.UpgradeRequest": {
"type": "object",
"properties": {
"candidateDigest": {
"description": "CandidateDigest is the digest of the candidate the check resolved.",
"type": "string",
"example": "sha256:fedcba9876543210fedcba9876543210fedcba9876543210fedcba9876543210"
},
"expectedDigest": {
"description": "ExpectedDigest is the digest of the installed package the caller saw.",
"type": "string",
"example": "sha256:0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef"
},
"generation": {
"description": "Generation is the installed generation the caller saw, when it knows it.",
"type": "integer",
"example": 1
},
"operationId": {
"description": "OperationID is the identity of the upgrade: repeat it to replay safely\nand, after an outcome_unknown answer, poll operation-status with it. The\nsame transition under another operation ID is refused.",
"type": "string",
"example": "upgrade-notes-1"
}
}
},
"apps.AdminView": {
"type": "object",
"properties": {
"canStart": {
"type": "boolean"
},
"canStop": {
"type": "boolean"
},
"managementUrl": {
"description": "ManagementURL opens the management view of the addon.",
"type": "string",
"example": "/app/addons?manage=notes"
}
}
},
"apps.App": {
"type": "object",
"properties": {
"admin": {
"description": "Admin is present only for administrators.",
"allOf": [
{
"$ref": "#/definitions/apps.AdminView"
}
]
},
"description": {
"type": "string",
"example": "Keep notes."
},
"descriptions": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"enabled": {
"description": "Enabled is the durable desired state. Health is State, a separate\nobservation: an enabled app can be failed.",
"type": "boolean"
},
"generation": {
"description": "Generation is the installed generation. A Start or Stop request must\ncarry the value that was listed; a stale value is denied.",
"type": "integer",
"example": 3
},
"hasUI": {
"description": "HasUI is true when the addon contributes any UI.",
"type": "boolean"
},
"icon": {
"description": "Icon is a same-origin URL of the verified immutable icon of this\ngeneration. It is absent when the manifest declares none.",
"type": "string",
"example": "/api/addons/apps/notes/icon?generation=3"
},
"id": {
"description": "ID is the addon identifier and Name its stable machine name. They are\nthe same in version one: the receipt records no separate catalog name.",
"type": "string",
"example": "notes"
},
"launchMode": {
"type": "string",
"example": "always"
},
"mainPage": {
"description": "MainPage is the path the Open action navigates to. It is present only\nfor an enabled addon whose manifest declares a main page that is one of\nits pages.",
"type": "string",
"example": "/app/addons/notes/home"
},
"name": {
"type": "string",
"example": "notes"
},
"state": {
"description": "State is running, starting, failed or stopped: the observation of the\nprocess, never an update.",
"type": "string",
"enum": [
"running",
"starting",
"failed",
"stopped"
],
"example": "running"
},
"title": {
"description": "Title and Description are the default-language text of the manifest.\nTitles and Descriptions carry every language when the manifest\nlocalizes the text, keyed by BCP 47 tag; they are absent otherwise.",
"type": "string",
"example": "Notes"
},
"titles": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"toolCount": {
"type": "integer"
},
"toolDefinitions": {
"description": "ToolDefinitions are the model-facing definitions the manifest declares:\nthe argument schema and the kind of each tool. The host serves them to\nchat from the verified manifest; the add-on process is never asked.",
"type": "array",
"items": {
"$ref": "#/definitions/contract.ToolDefinition"
}
},
"tools": {
"description": "Tools are the tool names of the manifest, at most 32. ToolCount is the\nfull number. The manifest carries no tool descriptions.",
"type": "array",
"items": {
"type": "string"
}
},
"ui": {
"description": "UI describes what the shell loads for this app. It is present only for\nan enabled addon with a UI whose installed package has a sha256 digest.",
"allOf": [
{
"$ref": "#/definitions/apps.UIView"
}
]
},
"update": {
"description": "Update is update-pending when the owner's cached update check says an\nupdate is available for the installed package. It is a separate\nobservation from State and is absent otherwise: an unchecked, failed\nor up to date add-on has none. The list never reaches a library.",
"type": "string",
"enum": [
"update-pending"
],
"example": "update-pending"
},
"version": {
"type": "string",
"example": "1.2.3"
}
}
},
"apps.ChatActionView": {
"type": "object",
"properties": {
"component": {
"type": "string",
"example": "SummaryPanel"
},
"icon": {
"description": "Icon is a host icon name, or the empty string when the action uses the\npackaged icon of the app (see the icon of the app).",
"type": "string",
"example": "sparkles"
},
"id": {
"description": "ID is the manifest identifier of the action.",
"type": "string",
"example": "summarize"
},
"label": {
"description": "Label is the default-language text; Labels carries every language when\nthe manifest localizes it.",
"type": "string",
"example": "Summarize"
},
"labels": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"packagedIcon": {
"description": "PackagedIcon is true when the action draws the packaged app icon.",
"type": "boolean"
},
"placement": {
"description": "Placement is composer or message.",
"type": "string",
"enum": [
"composer",
"message"
],
"example": "composer"
},
"tool": {
"description": "Tool is the own tool the action calls, or Component the exported\ncomponent it opens. Exactly one is set.",
"type": "string",
"example": "summarize"
}
}
},
"apps.List": {
"type": "object",
"properties": {
"apps": {
"type": "array",
"items": {
"$ref": "#/definitions/apps.App"
}
}
}
},
"apps.SlotView": {
"type": "object",
"properties": {
"component": {
"type": "string",
"example": "NotesSettings"
},
"id": {
"type": "string",
"example": "notes-nav"
},
"slot": {
"type": "string",
"enum": [
"chat.composer.actions",
"chat.message.actions",
"app.nav",
"settings.sections"
],
"example": "settings.sections"
}
}
},
"apps.StartRequest": {
"type": "object",
"properties": {
"generation": {
"description": "Generation is the installed generation the caller saw in the list.",
"type": "integer",
"example": 3
},
"operationId": {
"description": "OperationID names the lifecycle operation. Repeating it never runs the\nchange twice and, after an outcome_unknown answer, it is the identifier\nto poll.",
"type": "string",
"example": "start-notes-1"
}
}
},
"apps.UIView": {
"type": "object",
"properties": {
"actions": {
"description": "Actions are the chat actions that declare a placement and a target.\nSlots and Renderers are the other declared contributions. The manifest\nwas validated at install: every tool named here is a tool of the addon,\nevery slot is a host slot and every icon is a host icon name or the\npackaged icon. The host draws them only while the addon is enabled and\nusable by the caller.",
"type": "array",
"items": {
"$ref": "#/definitions/apps.ChatActionView"
}
},
"chatActions": {
"type": "array",
"items": {
"type": "string"
}
},
"css": {
"description": "CSS are the URLs of the declared style sheets.",
"type": "array",
"items": {
"type": "string"
}
},
"entry": {
"description": "Entry is the URL of the packaged ES module. Every asset URL names the\npackage digest, so a new package is a new URL.",
"type": "string",
"example": "/api/addons/notes/assets/0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef/ui.js"
},
"navigation": {
"type": "array",
"items": {
"type": "string"
}
},
"pages": {
"description": "Pages, Navigation and ChatActions are the identifiers the manifest\ndeclares. The host decides where each one is shown.",
"type": "array",
"items": {
"type": "string"
}
},
"renderers": {
"type": "array",
"items": {
"$ref": "#/definitions/contract.Renderer"
}
},
"sdk": {
"description": "SDK is the range of host SDK versions the module accepts.",
"allOf": [
{
"$ref": "#/definitions/contract.VersionRange"
}
]
},
"slots": {
"type": "array",
"items": {
"$ref": "#/definitions/apps.SlotView"
}
}
}
},
"auth.WebSocketTicketResponse": {
"type": "object",
"properties": {
"expires_at": {
"type": "string"
},
"ticket": {
"type": "string"
}
}
},
"backendlib.Update": {
"type": "object",
"properties": {
"digest": {
"type": "string"
},
"installedDigest": {
"type": "string"
},
"installedVersion": {
"description": "InstalledVersion and InstalledDigest are what metadata.json pins.",
"type": "string"
},
"name": {
"description": "Name is the backend name.",
"type": "string"
},
"signed": {
"description": "Signed says the library requires a signature, so the install of the\nupdate has a verifying step.",
"type": "boolean"
},
"sourceId": {
"description": "SourceID is the library the backend came from.",
"type": "string"
},
"version": {
"description": "Version and Digest are the build the library lists now. Version is what\nan install of the update names.",
"type": "string"
}
}
},
"builtin.Entry": {
"type": "object",
"properties": {
"adminOnly": {
"type": "boolean"
},
"feature": {
"type": "string"
},
"icon": {
"type": "string"
},
"id": {
"type": "string"
},
"kind": {
"$ref": "#/definitions/builtin.Kind"
},
"mainPath": {
"type": "string"
},
"titleKey": {
"type": "string"
}
}
},
"builtin.Kind": {
"type": "string",
"enum": [
"page"
],
"x-enum-varnames": [
"KindPage"
]
},
"catalog.Kind": {
"type": "string",
"enum": [
"addon",
"backend",
"model",
"all"
],
"x-enum-varnames": [
"KindAddon",
"KindBackend",
"KindModel",
"QueryAllKinds"
]
},
"catalog.RefreshState": {
"type": "object",
"properties": {
"checkedAt": {
"type": "string"
},
"count": {
"type": "integer"
},
"failed": {
"description": "Failed is true when the latest fetch failed; the previous releases are\nnot served in that case.",
"type": "boolean"
},
"lastSuccess": {
"type": "string"
},
"skipped": {
"description": "Skipped and SkippedKinds are as in RefreshResult.",
"type": "integer"
},
"skippedKinds": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"catalog.VerificationPolicy": {
"type": "object",
"properties": {
"identity": {
"type": "string"
},
"identityRegex": {
"description": "IdentityRegex is an anchored pattern for the signer identity, used\ninstead of Identity so that master and tagged builds both verify. Only a\nlibrary that serves backends may use it; an add-on install refuses it.",
"type": "string"
},
"issuer": {
"type": "string"
},
"requireSignature": {
"type": "boolean"
}
}
},
"config.Gallery": {
"type": "object",
"properties": {
"artifact_verification": {
"description": "ArtifactVerification overrides Verification only for the gallery OCI artifact.\nBackend images keep their separate Verification policy.",
"allOf": [
{
"$ref": "#/definitions/config.GalleryVerification"
}
]
},
"library": {
"description": "Library says the URL is a model library (the shared library entry\nshape), not a gallery index.",
"type": "boolean"
},
"mirrors": {
"description": "Mirrors are tried in order when URL cannot be fetched. They are a\nfallback for availability, not a load-balancing pool: the primary is\nalways preferred, and a mirror is only consulted after the one before\nit fails. Any URI the gallery loader understands works here\n(https://, github:, file://, oci://).",
"type": "array",
"items": {
"type": "string"
}
},
"name": {
"type": "string"
},
"url": {
"type": "string"
},
"verification": {
"$ref": "#/definitions/config.GalleryVerification"
}
}
},
"config.GalleryVerification": {
"type": "object",
"properties": {
"identity": {
"type": "string"
},
"identity_regex": {
"type": "string"
},
"issuer": {
"type": "string"
},
"issuer_regex": {
"type": "string"
},
"not_before": {
"description": "NotBefore is an RFC3339 timestamp. Empty disables the time check.",
"type": "string"
},
"source_repository": {
"description": "SourceRepository is an https URL compared exactly against the\ncertificate's source-repository extension. Empty skips the check.",
"type": "string"
}
}
},
"config.TTSVoice": {
"type": "object",
"properties": {
"gender": {
"type": "string"
},
"language": {
"type": "string"
},
"name": {
"type": "string"
}
}
},
"contract.OutputDescription": {
"type": "object",
"properties": {
"handle": {
"type": "string"
},
"mediaType": {
"type": "string"
},
"name": {
"type": "string"
},
"size": {
"type": "integer"
}
}
},
"contract.Receipt": {
"type": "object",
"properties": {
"operationId": {
"type": "string"
},
"result": {
"$ref": "#/definitions/contract.Result"
},
"revision": {
"$ref": "#/definitions/contract.Revision"
},
"state": {
"$ref": "#/definitions/contract.ReceiptState"
}
}
},
"contract.ReceiptState": {
"type": "string",
"enum": [
"pending",
"committed",
"rejected",
"unknown"
],
"x-enum-varnames": [
"ReceiptPending",
"ReceiptCommitted",
"ReceiptRejected",
"ReceiptUnknown"
]
},
"contract.Renderer": {
"type": "object",
"properties": {
"component": {
"type": "string"
},
"kind": {
"type": "string"
},
"tool": {
"type": "string"
}
}
},
"contract.Request": {
"type": "object",
"properties": {
"action": {
"type": "string"
},
"addonId": {
"type": "string"
},
"arguments": {
"type": "object"
},
"generation": {
"type": "integer"
},
"operationId": {
"type": "string"
},
"resourceId": {
"type": "string"
}
}
},
"contract.Result": {
"type": "object",
"properties": {
"outputs": {
"type": "array",
"items": {
"$ref": "#/definitions/contract.OutputDescription"
}
},
"payload": {
"type": "object"
}
}
},
"contract.Revision": {
"type": "object",
"properties": {
"data": {
"type": "integer"
},
"schema": {
"type": "integer"
}
}
},
"contract.ToolDefinition": {
"type": "object",
"properties": {
"description": {
"description": "Description is plain text for the model.",
"type": "string"
},
"inputSchema": {
"description": "InputSchema is the JSON schema of the arguments object. It is an\nobject schema with explicit properties, additionalProperties false at\nthe top level and the keywords the host allows only.",
"type": "object"
},
"kind": {
"description": "Kind is read, write, schema or create.",
"type": "string"
},
"name": {
"description": "Name is one of capabilities.tools.",
"type": "string"
}
}
},
"contract.VersionRange": {
"type": "object",
"properties": {
"max": {
"type": "integer"
},
"min": {
"type": "integer"
}
}
},
"failover.ChainState": {
"type": "string",
"enum": [
"primary",
"fallback",
"degraded"
],
"x-enum-varnames": [
"ChainPrimary",
"ChainFallback",
"ChainDegraded"
]
},
"failover.ChainStatus": {
"type": "object",
"properties": {
"active": {
"type": "string"
},
"active_since": {
"type": "string"
},
"name": {
"type": "string"
},
"pinned": {
"type": "string"
},
"state": {
"$ref": "#/definitions/failover.ChainState"
},
"targets": {
"type": "array",
"items": {
"$ref": "#/definitions/failover.TargetStatus"
}
}
}
},
"failover.Kind": {
"type": "string",
"enum": [
"local",
"remote"
],
"x-enum-varnames": [
"KindLocal",
"KindRemote"
]
},
"failover.TargetState": {
"type": "string",
"enum": [
"healthy",
"down",
"recovering",
"missing"
],
"x-enum-varnames": [
"StateHealthy",
"StateDown",
"StateRecovering",
"StateMissing"
]
},
"failover.TargetStatus": {
"type": "object",
"properties": {
"consecutive_ok": {
"type": "integer"
},
"kind": {
"$ref": "#/definitions/failover.Kind"
},
"last_error": {
"type": "string"
},
"last_probe": {
"type": "string"
},
"model": {
"type": "string"
},
"state": {
"$ref": "#/definitions/failover.TargetState"
},
"warm": {
"type": "boolean"
}
}
},
"functions.Function": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"name": {
"type": "string"
},
"parameters": {
"type": "object",
"additionalProperties": {}
},
"strict": {
"type": "boolean"
}
}
},
"functions.Item": {
"type": "object",
"properties": {
"properties": {
"type": "object",
"additionalProperties": {}
},
"type": {
"type": "string"
}
}
},
"functions.JSONFunctionStructure": {
"type": "object",
"properties": {
"$defs": {
"type": "object",
"additionalProperties": {}
},
"anyOf": {
"type": "array",
"items": {
"$ref": "#/definitions/functions.Item"
}
},
"oneOf": {
"type": "array",
"items": {
"$ref": "#/definitions/functions.Item"
}
}
}
},
"functions.Tool": {
"type": "object",
"properties": {
"function": {
"$ref": "#/definitions/functions.Function"
},
"type": {
"type": "string"
}
}
},
"gallery.File": {
"type": "object",
"properties": {
"filename": {
"type": "string"
},
"sha256": {
"type": "string"
},
"uri": {
"type": "string"
}
}
},
"gallery.GalleryBackend": {
"type": "object",
"properties": {
"alias": {
"type": "string"
},
"backend": {
"description": "Backend is the resolved backend engine for this model (e.g. \"llama-cpp\").\nPopulated at load time from overrides, inline config, or the URL-referenced config file.",
"type": "string"
},
"capabilities": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"description": {
"type": "string"
},
"files": {
"description": "AdditionalFiles are used to add additional files to the model",
"type": "array",
"items": {
"$ref": "#/definitions/gallery.File"
}
},
"gallery": {
"description": "Gallery is a reference to the gallery which contains the model",
"allOf": [
{
"$ref": "#/definitions/config.Gallery"
}
]
},
"icon": {
"type": "string"
},
"installed": {
"description": "Installed is used to indicate if the model is installed or not",
"type": "boolean"
},
"license": {
"type": "string"
},
"mirrors": {
"type": "array",
"items": {
"type": "string"
}
},
"name": {
"type": "string"
},
"sha256": {
"description": "SHA256 is the expected sha256 of the backend tarball at URI / Mirrors.\nEmpty disables the integrity check; OCI URIs carry their own digest.",
"type": "string"
},
"size": {
"description": "Size is an optional hardcoded model size string (e.g. \"500MB\", \"14.5GB\").\nUsed when the size cannot be estimated automatically.",
"type": "string"
},
"tags": {
"type": "array",
"items": {
"type": "string"
}
},
"uri": {
"type": "string"
},
"url": {
"type": "string"
},
"urls": {
"type": "array",
"items": {
"type": "string"
}
},
"version": {
"type": "string"
}
}
},
"gallery.Metadata": {
"type": "object",
"properties": {
"backend": {
"description": "Backend is the resolved backend engine for this model (e.g. \"llama-cpp\").\nPopulated at load time from overrides, inline config, or the URL-referenced config file.",
"type": "string"
},
"description": {
"type": "string"
},
"files": {
"description": "AdditionalFiles are used to add additional files to the model",
"type": "array",
"items": {
"$ref": "#/definitions/gallery.File"
}
},
"gallery": {
"description": "Gallery is a reference to the gallery which contains the model",
"allOf": [
{
"$ref": "#/definitions/config.Gallery"
}
]
},
"icon": {
"type": "string"
},
"installed": {
"description": "Installed is used to indicate if the model is installed or not",
"type": "boolean"
},
"license": {
"type": "string"
},
"name": {
"type": "string"
},
"size": {
"description": "Size is an optional hardcoded model size string (e.g. \"500MB\", \"14.5GB\").\nUsed when the size cannot be estimated automatically.",
"type": "string"
},
"tags": {
"type": "array",
"items": {
"type": "string"
}
},
"url": {
"type": "string"
},
"urls": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"gallery.ModelStorage": {
"type": "object",
"properties": {
"files": {
"type": "array",
"items": {
"type": "string"
}
},
"missing": {
"type": "array",
"items": {
"type": "string"
}
},
"name": {
"type": "string"
},
"shared_bytes": {
"type": "integer"
},
"size_bytes": {
"type": "integer"
}
}
},
"gallery.NodeDriftInfo": {
"type": "object",
"properties": {
"digest": {
"type": "string"
},
"node_id": {
"type": "string"
},
"node_name": {
"type": "string"
},
"version": {
"type": "string"
}
}
},
"gallery.StorageFile": {
"type": "object",
"properties": {
"missing": {
"type": "boolean"
},
"models": {
"type": "array",
"items": {
"type": "string"
}
},
"path": {
"description": "Path is relative to the models directory.",
"type": "string"
},
"size_bytes": {
"type": "integer"
}
}
},
"gallery.StorageIndex": {
"type": "object",
"properties": {
"errors": {
"description": "Errors carries per-config failures (malformed YAML, path escapes),\nreported rather than fatal so one broken config cannot blank the\nwhole report.",
"type": "array",
"items": {
"type": "string"
}
},
"files": {
"type": "array",
"items": {
"$ref": "#/definitions/gallery.StorageFile"
}
},
"models": {
"type": "array",
"items": {
"$ref": "#/definitions/gallery.ModelStorage"
}
},
"total_bytes": {
"type": "integer"
}
}
},
"gallery.UpgradeInfo": {
"type": "object",
"properties": {
"available_digest": {
"type": "string"
},
"available_version": {
"type": "string"
},
"backend_name": {
"type": "string"
},
"installed_digest": {
"type": "string"
},
"installed_version": {
"type": "string"
},
"node_drift": {
"description": "NodeDrift lists nodes whose installed version or digest differs from\nthe cluster majority. Non-empty means the cluster has diverged and an\nupgrade will realign it. Empty in single-node mode.",
"type": "array",
"items": {
"$ref": "#/definitions/gallery.NodeDriftInfo"
}
},
"reason": {
"description": "Reason says why the backend is flagged when version and digest alone\nwould not flag it: the backend failed a request in a way that shows it\nis older than the server. Empty for ordinary version or digest upgrades.",
"type": "string"
}
}
},
"gallery.Variant": {
"type": "object",
"properties": {
"model": {
"description": "Model is the name of a gallery entry that declares no variants of its own.",
"type": "string"
}
}
},
"galleryop.NodeProgress": {
"type": "object",
"properties": {
"current": {
"type": "string"
},
"error": {
"type": "string"
},
"file_name": {
"type": "string"
},
"node_id": {
"type": "string"
},
"node_name": {
"type": "string"
},
"percentage": {
"type": "number"
},
"phase": {
"type": "string"
},
"status": {
"type": "string"
},
"total": {
"type": "string"
}
}
},
"galleryop.OpStatus": {
"type": "object",
"properties": {
"cancellable": {
"description": "Cancellable is true if the operation can be cancelled",
"type": "boolean"
},
"cancelled": {
"description": "Cancelled is true if the operation was cancelled",
"type": "boolean"
},
"current_bytes": {
"type": "integer"
},
"deletion": {
"description": "Deletion is true if the operation is a deletion",
"type": "boolean"
},
"downloaded_size": {
"type": "string"
},
"file_name": {
"type": "string"
},
"file_size": {
"type": "string"
},
"gallery_element_name": {
"type": "string"
},
"message": {
"type": "string"
},
"nodes": {
"description": "Nodes is the per-node breakdown for a fanned-out backend install.\nPopulated by DistributedBackendManager (per-node terminal status)\nand by the Phase 2 progress bridge (per-byte ticks). The\n/api/operations handler surfaces this so the UI can render an\nexpandable per-node view of an in-flight install.",
"type": "array",
"items": {
"$ref": "#/definitions/galleryop.NodeProgress"
}
},
"phase": {
"type": "string"
},
"processed": {
"type": "boolean"
},
"progress": {
"type": "number"
},
"total_bytes": {
"type": "integer"
}
}
},
"gateway.ApplyResourceRequest": {
"type": "object",
"properties": {
"arguments": {
"description": "Arguments are the tool's own arguments without the resource, which the\npath names, and without the tool, which Tool names.",
"type": "object"
},
"generation": {
"description": "Generation is the installed generation the caller listed.",
"type": "integer",
"example": 3
},
"operationId": {
"description": "OperationID names the change. Repeating it with the same request\nreturns the same receipt and never applies twice; after an\noutcome_unknown answer it is what to repeat.",
"type": "string",
"example": "apply-1"
},
"revision": {
"description": "Revision is the revision the caller read. A moved revision is a stale\ndenial and nothing is changed.",
"allOf": [
{
"$ref": "#/definitions/contract.Revision"
}
]
},
"tool": {
"description": "Tool is a write or schema tool the add-on declares.",
"type": "string",
"example": "rows-insert"
}
}
},
"gateway.ApplyResult": {
"type": "object",
"properties": {
"operationId": {
"type": "string"
},
"requestHash": {
"description": "RequestHash binds the operation ID to the request: the SHA-256 of the\nresource, the tool and the canonical arguments. A lookup of the\noperation reports it, so a client can tell its own request from\nanother one that used the same ID.",
"type": "string",
"example": "3f786850e387550fdab836ed7e6dc881de23001b3f786850e387550fdab836ed"
},
"resourceId": {
"type": "string"
},
"result": {
"description": "Result is what the add-on's receipt reported, for example the count\nand the identifiers of the changed rows.",
"type": "object"
},
"revision": {
"description": "Revision is the revision after the change.",
"allOf": [
{
"$ref": "#/definitions/contract.Revision"
}
]
}
}
},
"gateway.Artifact": {
"type": "object",
"properties": {
"addonId": {
"description": "AddonID and Generation name the addon build that produced the result.",
"type": "string"
},
"createdAt": {
"type": "string"
},
"expiresAt": {
"description": "ExpiresAt is when the artifact stops being served and is removed.",
"type": "string"
},
"generation": {
"type": "integer"
},
"id": {
"description": "ID is the host-generated opaque identifier. It is the only handle a\nclient ever holds for the file.",
"type": "string"
},
"mediaType": {
"type": "string"
},
"name": {
"description": "Name is the sanitized attachment name and MediaType the safe media type\nthe download is served with.",
"type": "string"
},
"operationId": {
"description": "OperationID is the operation of the request that produced the output.",
"type": "string"
},
"principal": {
"description": "Principal is the user who requested the result. Only this user can read\nor download the artifact.",
"type": "string"
},
"reason": {
"description": "Reason says why a transfer failed or was cancelled. It never carries\naddon text.",
"type": "string"
},
"size": {
"description": "Size is the exact number of bytes of a complete artifact.",
"type": "integer"
},
"sources": {
"description": "Sources are the resources the bytes were produced from. Every one of\nthem must still be readable by the principal at completion and at each\ndownload.",
"type": "array",
"items": {
"$ref": "#/definitions/gateway.ArtifactResource"
}
},
"state": {
"$ref": "#/definitions/gateway.ArtifactState"
}
}
},
"gateway.ArtifactResource": {
"type": "object",
"properties": {
"addonId": {
"type": "string"
},
"resourceId": {
"type": "string"
}
}
},
"gateway.ArtifactState": {
"type": "string",
"enum": [
"transferring",
"complete",
"failed",
"cancelled"
],
"x-enum-varnames": [
"ArtifactTransferring",
"ArtifactComplete",
"ArtifactFailed",
"ArtifactCancelled"
]
},
"gateway.BlockingProposal": {
"type": "object",
"properties": {
"action": {
"type": "string"
},
"addonId": {
"type": "string"
},
"ageSeconds": {
"description": "AgeSeconds is how long the outcome has been unresolved, counted from\nthe hand-over to the addon.",
"type": "integer"
},
"createdAt": {
"type": "string"
},
"digest": {
"description": "Digest is what the owner names to resolve the proposal.",
"type": "string"
},
"effects": {
"$ref": "#/definitions/gateway.ProposalEffects"
},
"expiresAt": {
"type": "string"
},
"id": {
"description": "ID is the proposal identifier, which is its operation identifier.",
"type": "string"
},
"proposerId": {
"description": "ProposerID is the user who proposed the operation.",
"type": "string"
},
"reason": {
"description": "Reason says why the outcome is unresolved.",
"type": "string"
},
"resolution": {
"description": "Resolution is present once the owner stopped waiting.",
"allOf": [
{
"$ref": "#/definitions/gateway.ProposalResolution"
}
]
},
"resourceId": {
"type": "string"
},
"state": {
"description": "State is executing, outcome-unknown or abandoned-unknown.",
"allOf": [
{
"$ref": "#/definitions/gateway.ProposalState"
}
]
}
}
},
"gateway.CreateResourceRequest": {
"type": "object",
"properties": {
"addonId": {
"description": "AddonID names the add-on whose resource is created.",
"type": "string",
"example": "tables"
},
"arguments": {
"description": "Arguments are the create arguments the add-on's create tool takes, for\nexample the name. They are validated against its tool definition.",
"type": "object"
},
"generation": {
"description": "Generation is the installed generation the caller listed.",
"type": "integer",
"example": 3
},
"operationId": {
"description": "OperationID names the creation. Repeating it never creates a second\nresource and, after an outcome_unknown answer, it is what to repeat.",
"type": "string",
"example": "create-budget-1"
}
}
},
"gateway.DeleteResourceRequest": {
"type": "object",
"properties": {
"confirm": {
"description": "Confirm must equal the display name of the resource exactly: the person\ntypes it to confirm. Deletion removes the data and every grant, cannot\nbe undone and is never a chat operation.",
"type": "string",
"example": "Budget"
},
"generation": {
"type": "integer",
"example": 3
},
"operationId": {
"type": "string",
"example": "delete-1"
},
"revision": {
"$ref": "#/definitions/contract.Revision"
}
}
},
"gateway.DeleteResult": {
"type": "object",
"properties": {
"deleted": {
"description": "Deleted is true once the add-on confirmed that the data is gone and\nthe grants were removed.",
"type": "boolean"
},
"resourceId": {
"type": "string"
}
}
},
"gateway.Grant": {
"type": "object",
"properties": {
"displayName": {
"type": "string",
"example": "Bob"
},
"role": {
"type": "string",
"enum": [
"viewer",
"editor"
],
"example": "viewer"
},
"userId": {
"description": "UserID is the immutable account identifier of the grantee.",
"type": "string",
"example": "4f1c"
}
}
},
"gateway.GrantList": {
"type": "object",
"properties": {
"grants": {
"type": "array",
"items": {
"$ref": "#/definitions/gateway.Grant"
}
},
"resourceId": {
"type": "string"
}
}
},
"gateway.OperationStatus": {
"type": "object",
"properties": {
"operationId": {
"type": "string"
},
"requestHash": {
"type": "string",
"example": "3f786850e387550fdab836ed7e6dc881de23001b3f786850e387550fdab836ed"
},
"state": {
"description": "State is pending, committed, rejected or unknown.",
"type": "string",
"enum": [
"pending",
"committed",
"rejected",
"unknown"
]
},
"tool": {
"description": "Tool and RequestHash name the request the operation ID was first\nused for, as the host recorded it when it dispatched the change (kind\napply only). They are absent when the host no longer holds the\nrequest, for example after a restart. Compare RequestHash with the\nhash of your own request before you believe a committed state.",
"type": "string",
"example": "rows-insert"
}
}
},
"gateway.PendingList": {
"type": "object",
"properties": {
"proposals": {
"description": "Proposals are the caller's own proposals that still wait for a human\ndecision, newest first. A proposal of another user is never listed.",
"type": "array",
"items": {
"$ref": "#/definitions/gateway.Proposal"
}
}
}
},
"gateway.Proposal": {
"type": "object",
"properties": {
"action": {
"type": "string"
},
"addonId": {
"type": "string"
},
"arguments": {
"description": "Arguments are the canonical stored arguments. They are the only\narguments that can ever be submitted for this proposal.",
"type": "object"
},
"blocked": {
"description": "Blocked reports that another operation on the resource has an\nunresolved outcome, so confirming is denied until it is resolved. The\noperation itself belongs to its proposer and is shown only to the\nresource's owner role, in the unresolved listing and in the denial.",
"type": "boolean"
},
"createdAt": {
"type": "string"
},
"digest": {
"description": "Digest binds a confirmation to exactly what was shown.",
"type": "string"
},
"effects": {
"$ref": "#/definitions/gateway.ProposalEffects"
},
"expiresAt": {
"type": "string"
},
"generation": {
"type": "integer"
},
"id": {
"description": "ID is the proposal identifier. It is the operation identifier of the\nstored operation, so a lost reply is always resolved by the same ID.",
"type": "string"
},
"operationId": {
"type": "string"
},
"reason": {
"description": "Reason says why a proposal is cancelled, expired or rejected.",
"type": "string"
},
"receipt": {
"description": "Receipt is present once the runtime answered.",
"allOf": [
{
"$ref": "#/definitions/contract.Receipt"
}
]
},
"resolution": {
"description": "Resolution is present once the resource owner stopped waiting for an\nunresolved outcome. It stays after a late receipt is recorded.",
"allOf": [
{
"$ref": "#/definitions/gateway.ProposalResolution"
}
]
},
"resourceId": {
"type": "string"
},
"revision": {
"description": "Revision is the schema and data revision the preview saw. The runtime\nrefuses the apply when it moved.",
"allOf": [
{
"$ref": "#/definitions/contract.Revision"
}
]
},
"state": {
"$ref": "#/definitions/gateway.ProposalState"
}
}
},
"gateway.ProposalEffects": {
"type": "object",
"properties": {
"count": {
"description": "Count is the exact number of effects the operation would have.",
"type": "integer"
},
"sample": {
"description": "Sample is a bounded excerpt of the effects, or the whole preview summary\nwhen the addon did not separate one.",
"type": "object"
}
}
},
"gateway.ProposalList": {
"type": "object",
"properties": {
"proposals": {
"type": "array",
"items": {
"$ref": "#/definitions/gateway.BlockingProposal"
}
}
}
},
"gateway.ProposalResolution": {
"type": "object",
"properties": {
"actorId": {
"description": "ActorID is the resource owner who acknowledged the unknown outcome.",
"type": "string"
},
"at": {
"description": "At is when the acknowledgement was recorded.",
"type": "string"
},
"reason": {
"description": "Reason is the owner's stated reason.",
"type": "string"
}
}
},
"gateway.ProposalState": {
"type": "string",
"enum": [
"pending",
"cancelled",
"expired",
"executing",
"applied",
"rejected",
"outcome-unknown",
"abandoned-unknown"
],
"x-enum-varnames": [
"ProposalPending",
"ProposalCancelled",
"ProposalExpired",
"ProposalExecuting",
"ProposalApplied",
"ProposalRejected",
"ProposalOutcomeUnknown",
"ProposalAbandoned"
]
},
"gateway.RelayRequest": {
"type": "object",
"properties": {
"digest": {
"description": "Digest names the installed package an asset belongs to: the lower-case\nhex of its sha256. Only the kind \"asset\" uses it.",
"type": "string"
},
"handle": {
"type": "string"
},
"kind": {
"type": "string"
},
"request": {
"$ref": "#/definitions/contract.Request"
}
}
},
"gateway.RenameResourceRequest": {
"type": "object",
"properties": {
"generation": {
"type": "integer",
"example": 3
},
"name": {
"type": "string",
"example": "Budget 2027"
},
"operationId": {
"type": "string",
"example": "rename-1"
},
"revision": {
"description": "Revision is the revision the caller listed. A moved revision is a\nstale denial and nothing is changed.",
"allOf": [
{
"$ref": "#/definitions/contract.Revision"
}
]
}
}
},
"gateway.ResourceInfo": {
"type": "object",
"properties": {
"addonId": {
"type": "string",
"example": "tables"
},
"id": {
"description": "ID is the opaque resource identifier.",
"type": "string",
"example": "r0123456789abcdef0123456789abcde"
},
"name": {
"description": "Name is the display name the add-on reported. It is empty when the\nadd-on could not answer in time.",
"type": "string",
"example": "Budget"
},
"owner": {
"description": "Owner is the display name of the owner, shown to people the resource\nis shared with. It is empty in local-owner mode.",
"type": "string",
"example": "Alice"
},
"revision": {
"description": "Revision is the schema and data revision the listing saw. It is\npresent for the owner role and goes back with a rename or a delete.",
"allOf": [
{
"$ref": "#/definitions/contract.Revision"
}
]
},
"role": {
"description": "Role is the caller's own role: viewer, editor or owner.",
"type": "string",
"enum": [
"viewer",
"editor",
"owner"
],
"example": "owner"
},
"sharedWith": {
"description": "SharedWith is the number of people the resource is shared with. It is\npresent for the owner role only.",
"type": "integer",
"example": 2
}
}
},
"gateway.ResourceList": {
"type": "object",
"properties": {
"addonId": {
"type": "string",
"example": "tables"
},
"generation": {
"description": "Generation is the installed generation the listing saw. Rename, delete\nand create carry it; a stale tab is denied.",
"type": "integer",
"example": 3
},
"noun": {
"description": "Noun is the singular word the add-on uses for its resources.",
"type": "string",
"example": "database"
},
"operations": {
"description": "Operations lists the management operations the add-on supports.",
"type": "array",
"items": {
"type": "string",
"enum": [
"create",
"rename",
"delete"
]
}
},
"resources": {
"type": "array",
"items": {
"$ref": "#/definitions/gateway.ResourceInfo"
}
},
"sharing": {
"description": "Sharing is false in local-owner mode, where nothing can be shared.",
"type": "boolean"
},
"truncated": {
"description": "Truncated says more resources exist than the bound lists.",
"type": "boolean"
}
}
},
"gateway.Selection": {
"type": "object",
"properties": {
"addonId": {
"type": "string",
"example": "tables"
},
"conversationId": {
"type": "string",
"example": "chat-1"
},
"name": {
"type": "string",
"example": "Budget"
},
"resourceId": {
"type": "string",
"example": "r0123456789abcdef0123456789abcde"
},
"role": {
"type": "string",
"enum": [
"viewer",
"editor",
"owner"
]
}
}
},
"gateway.SelectionRequest": {
"type": "object",
"properties": {
"resourceId": {
"description": "ResourceID is the resource to use in the conversation. The caller must\nhold at least the viewer role on it now.",
"type": "string",
"example": "r0123456789abcdef0123456789abcde"
}
}
},
"gateway.SetGrantRequest": {
"type": "object",
"properties": {
"operationId": {
"description": "OperationID is optional. It is returned with outcome_unknown, and the\nchange is safe to repeat.",
"type": "string",
"example": "grant-1"
},
"role": {
"description": "Role is viewer or editor. The owner role cannot be granted.",
"type": "string",
"enum": [
"viewer",
"editor"
],
"example": "viewer"
}
}
},
"gateway.UserList": {
"type": "object",
"properties": {
"users": {
"type": "array",
"items": {
"$ref": "#/definitions/gateway.UserRef"
}
}
}
},
"gateway.UserRef": {
"type": "object",
"properties": {
"displayName": {
"type": "string",
"example": "Bob"
},
"id": {
"type": "string",
"example": "4f1c"
}
}
},
"library.Info": {
"type": "object",
"properties": {
"authReference": {
"type": "string"
},
"conflict": {
"description": "Conflict explains how this source clashes with another one, when it does.",
"type": "string"
},
"default": {
"type": "boolean"
},
"enabled": {
"description": "Enabled is what the host does with the source: false for a source that\nis turned off, removed, or a shipped default while the official\nlibraries are off. Suspended tells the last case apart.",
"type": "boolean"
},
"format": {
"type": "string"
},
"id": {
"type": "string"
},
"kind": {
"$ref": "#/definitions/catalog.Kind"
},
"mirrors": {
"description": "Mirrors are the fallback addresses of the source, in order.",
"type": "array",
"items": {
"type": "string"
}
},
"name": {
"type": "string"
},
"policy": {
"$ref": "#/definitions/catalog.VerificationPolicy"
},
"refresh": {
"$ref": "#/definitions/catalog.RefreshState"
},
"removed": {
"type": "boolean"
},
"suspended": {
"type": "boolean"
},
"url": {
"type": "string"
}
}
},
"localai.APIInstructionResponse": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"name": {
"type": "string"
},
"swagger_fragment": {
"type": "object",
"additionalProperties": {}
},
"tags": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"localai.AliasInfo": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"target": {
"type": "string"
}
}
},
"localai.BackendLibraryInstallRequest": {
"type": "object",
"properties": {
"name": {
"description": "Name is the backend name of the entry.",
"type": "string"
},
"sourceId": {
"description": "SourceID is the library the entry comes from.",
"type": "string"
},
"version": {
"description": "Version is the version or the channel of the entry.",
"type": "string"
}
}
},
"localai.BrandingResponse": {
"type": "object",
"properties": {
"favicon_url": {
"type": "string"
},
"instance_name": {
"type": "string"
},
"instance_tagline": {
"type": "string"
},
"logo_horizontal_url": {
"type": "string"
},
"logo_url": {
"type": "string"
}
}
},
"localai.FailoverChainsResponse": {
"type": "object",
"properties": {
"chains": {
"type": "array",
"items": {
"$ref": "#/definitions/failover.ChainStatus"
}
}
}
},
"localai.FailoverPinRequest": {
"type": "object",
"properties": {
"target": {
"type": "string"
}
}
},
"localai.GalleryBackend": {
"type": "object",
"properties": {
"force": {
"description": "Force reinstalls the backend even when it is already installed and\nrunnable. Off by default so apply stays idempotent for supervising\napps that ensure their backend on every boot.",
"type": "boolean"
},
"id": {
"type": "string"
}
}
},
"localai.GalleryModel": {
"type": "object",
"properties": {
"backend": {
"description": "Backend is the resolved backend engine for this model (e.g. \"llama-cpp\").\nPopulated at load time from overrides, inline config, or the URL-referenced config file.",
"type": "string"
},
"config_file": {
"description": "config_file is read in the situation where URL is blank - and therefore this is a base config.",
"type": "object",
"additionalProperties": {}
},
"description": {
"type": "string"
},
"files": {
"description": "AdditionalFiles are used to add additional files to the model",
"type": "array",
"items": {
"$ref": "#/definitions/gallery.File"
}
},
"gallery": {
"description": "Gallery is a reference to the gallery which contains the model",
"allOf": [
{
"$ref": "#/definitions/config.Gallery"
}
]
},
"icon": {
"type": "string"
},
"id": {
"type": "string"
},
"installed": {
"description": "Installed is used to indicate if the model is installed or not",
"type": "boolean"
},
"license": {
"type": "string"
},
"name": {
"type": "string"
},
"overrides": {
"description": "Overrides are used to override the configuration of the model located at URL",
"type": "object",
"additionalProperties": {}
},
"size": {
"description": "Size is an optional hardcoded model size string (e.g. \"500MB\", \"14.5GB\").\nUsed when the size cannot be estimated automatically.",
"type": "string"
},
"tags": {
"type": "array",
"items": {
"type": "string"
}
},
"url": {
"type": "string"
},
"urls": {
"type": "array",
"items": {
"type": "string"
}
},
"variant": {
"description": "Variant installs one specific build of an entry that declares variants,\nnamed as it appears in the entry's ` + "`" + `variants` + "`" + ` list (see the ` + "`" + `variants` + "`" + `\nand ` + "`" + `auto_variant` + "`" + ` fields of the gallery listing). Leave it empty to let\nLocalAI auto-select the largest build this host can actually run.",
"type": "string"
},
"variants": {
"description": "Variants is an optional, UNORDERED list of alternative builds of the same\nmodel (other backends such as MLX or vLLM, other quantizations) that the\ninstaller may pick instead of this entry's own payload. Authoring is\ndeliberately dumb: name the models, and the selector works out which one\nthis host should get.\n\nThe entry itself is always the last resort, so an entry carrying variants\nstays a complete, installable entry and older LocalAI releases, which drop\nthis key, install it exactly as before.",
"type": "array",
"items": {
"$ref": "#/definitions/gallery.Variant"
}
}
}
},
"localai.LibrarySourcesResponse": {
"type": "object",
"properties": {
"officialLibraries": {
"type": "boolean"
},
"sources": {
"type": "array",
"items": {
"$ref": "#/definitions/library.Info"
}
}
}
},
"localai.ModelExportsStatus": {
"type": "object",
"properties": {
"model_exports": {
"description": "ModelExports is false only when this instance does not see the models\npath of the add-on owner. It is the same value as the model_exports key\nof GET /api/features.",
"type": "boolean"
}
}
},
"localai.ModelResponse": {
"type": "object",
"properties": {
"config": {},
"config_revision": {
"type": "string"
},
"details": {
"type": "array",
"items": {
"type": "string"
}
},
"error": {
"type": "string"
},
"filename": {
"type": "string"
},
"message": {
"type": "string"
},
"pending_cleanup": {
"type": "integer"
},
"success": {
"type": "boolean"
}
}
},
"localai.MotionSessionInfo": {
"type": "object",
"properties": {
"frames": {
"type": "integer"
},
"id": {
"type": "string"
},
"model": {
"type": "string"
},
"poses": {
"type": "integer"
},
"profile": {
"type": "string"
},
"time_origin": {
"type": "string"
},
"upload_accepted": {
"type": "integer"
},
"upload_dropped": {
"type": "integer"
}
}
},
"localai.MotionSessionRequest": {
"type": "object",
"properties": {
"model": {
"type": "string"
},
"profile": {
"type": "string"
},
"time_origin": {
"type": "string"
}
}
},
"localai.MotionTicketRequest": {
"type": "object",
"properties": {
"origin": {
"type": "string"
}
}
},
"localai.TTSModelVoices": {
"type": "object",
"properties": {
"model": {
"type": "string"
},
"voices": {
"type": "array",
"items": {
"$ref": "#/definitions/config.TTSVoice"
}
}
}
},
"localai.TTSVoicesResponse": {
"type": "object",
"properties": {
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/localai.TTSModelVoices"
}
}
}
},
"localai.UpdateMaxReplicasPerModelRequest": {
"type": "object",
"properties": {
"value": {
"description": "Value is the new per-model replica cap on this node. Must be \u003e= 1.",
"type": "integer"
}
}
},
"localai.UpdateVRAMBudgetRequest": {
"type": "object",
"properties": {
"value": {
"description": "Value is the VRAM cap (\"80%\" or \"12GB\"). Empty string clears the cap.",
"type": "string"
}
}
},
"localai.VoiceProfileListResponse": {
"type": "object",
"properties": {
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/voiceprofile.Profile"
}
}
}
},
"middleware.TraceBucket": {
"type": "object",
"properties": {
"count": {
"type": "integer"
},
"errors": {
"type": "integer"
},
"start": {
"type": "string"
}
}
},
"middleware.TraceSummary": {
"type": "object",
"properties": {
"buckets": {
"type": "array",
"items": {
"$ref": "#/definitions/middleware.TraceBucket"
}
},
"errors": {
"type": "integer"
},
"p95_ms": {
"type": "integer"
},
"total": {
"type": "integer"
},
"window_hours": {
"type": "integer"
}
}
},
"model.BackendLogLine": {
"type": "object",
"properties": {
"stream": {
"description": "\"stdout\" or \"stderr\"",
"type": "string"
},
"text": {
"type": "string"
},
"timestamp": {
"type": "string"
}
}
},
"modeladmin.VRAMRequest": {
"type": "object",
"properties": {
"context_size": {
"type": "integer"
},
"gpu_layers": {
"type": "integer"
},
"kv_quant_bits": {
"type": "integer"
},
"model": {
"type": "string"
}
}
},
"modeladmin.VRAMResponse": {
"type": "object",
"properties": {
"context_length": {
"type": "integer"
},
"context_note": {
"type": "string"
},
"model_max_context": {
"type": "integer"
},
"size_bytes": {
"type": "integer"
},
"size_display": {
"type": "string"
},
"vram_bytes": {
"type": "integer"
},
"vram_display": {
"type": "string"
}
}
},
"nodes.NodeModel": {
"type": "object",
"properties": {
"address": {
"description": "gRPC address for this replica's backend process",
"type": "string"
},
"backend_type": {
"description": "e.g. \"llama-cpp\"; used by reconciler to replicate loads",
"type": "string"
},
"cleanup_attempts": {
"type": "integer"
},
"cleanup_error": {
"type": "string"
},
"cleanup_next_retry_at": {
"type": "string"
},
"config_revision": {
"type": "string"
},
"created_at": {
"type": "string"
},
"effective_options_hash": {
"type": "string"
},
"id": {
"type": "string"
},
"in_flight": {
"description": "number of active requests on this replica",
"type": "integer"
},
"last_used": {
"type": "string"
},
"loading_by": {
"description": "frontend ID that triggered loading",
"type": "string"
},
"model_name": {
"type": "string"
},
"node_id": {
"type": "string"
},
"replica_index": {
"type": "integer"
},
"state": {
"description": "staging, loading, loaded, unloading, idle",
"type": "string"
},
"updated_at": {
"type": "string"
}
}
},
"proto.MemoryUsageData": {
"type": "object",
"properties": {
"breakdown": {
"type": "object",
"additionalProperties": {
"type": "integer",
"format": "int64"
}
},
"total": {
"type": "integer"
}
}
},
"proto.SpeakerEncoder": {
"type": "object",
"properties": {
"dimension": {
"type": "integer"
},
"family": {
"description": "embedding space of the encoder; empty when the backend cannot tell",
"type": "string"
},
"identity": {
"description": "sha256 of loaded GGUF bytes",
"type": "string"
}
}
},
"proto.StatusResponse": {
"type": "object",
"properties": {
"memory": {
"$ref": "#/definitions/proto.MemoryUsageData"
},
"speaker_encoder": {
"description": "trusted metadata from the loaded server encoder, never request data",
"allOf": [
{
"$ref": "#/definitions/proto.SpeakerEncoder"
}
]
},
"state": {
"$ref": "#/definitions/proto.StatusResponse_State"
}
}
},
"proto.StatusResponse_State": {
"type": "integer",
"format": "int32",
"enum": [
0,
1,
2,
-1
],
"x-enum-varnames": [
"StatusResponse_UNINITIALIZED",
"StatusResponse_BUSY",
"StatusResponse_READY",
"StatusResponse_ERROR"
]
},
"proto.VADResponse": {
"type": "object",
"properties": {
"segments": {
"type": "array",
"items": {
"$ref": "#/definitions/proto.VADSegment"
}
}
}
},
"proto.VADSegment": {
"type": "object",
"properties": {
"end": {
"type": "number"
},
"start": {
"type": "number"
}
}
},
"recovery.BackupInfo": {
"type": "object",
"properties": {
"addonId": {
"type": "string"
},
"created": {
"type": "string"
},
"digest": {
"type": "string"
},
"files": {
"type": "integer"
},
"id": {
"type": "string"
},
"mode": {
"description": "Mode and Version are the ownership mode and package version it holds.",
"type": "string"
},
"pinned": {
"description": "Pinned is set while the backup is the recovery copy of an upgrade.",
"type": "boolean"
},
"reason": {
"type": "string"
},
"size": {
"type": "integer"
},
"verification": {
"type": "string"
},
"verified": {
"type": "boolean"
},
"version": {
"type": "string"
}
}
},
"recovery.BackupReceipt": {
"type": "object",
"properties": {
"addonId": {
"type": "string"
},
"created": {
"type": "string"
},
"digest": {
"type": "string"
},
"files": {
"type": "integer"
},
"id": {
"type": "string"
},
"mode": {
"description": "Mode and Version are the ownership mode and package version it holds.",
"type": "string"
},
"pinned": {
"description": "Pinned is set while the backup is the recovery copy of an upgrade.",
"type": "boolean"
},
"reason": {
"type": "string"
},
"size": {
"type": "integer"
},
"verified": {
"type": "boolean"
},
"version": {
"type": "string"
}
}
},
"recovery.PurgeReceipt": {
"type": "object",
"properties": {
"addonId": {
"type": "string"
},
"at": {
"type": "string"
},
"resources": {
"type": "integer"
},
"scope": {
"$ref": "#/definitions/recovery.PurgeScope"
},
"warnings": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"recovery.PurgeScope": {
"type": "object",
"properties": {
"backups": {
"description": "Backups are the backups of the add-on.",
"type": "boolean"
},
"data": {
"description": "Data is the add-on's data directory and the owner's records of its\nresources, grants and conversation selections.",
"type": "boolean"
},
"recoveryCopies": {
"description": "RecoveryCopies are the data copies that restores and failed upgrades\nset aside.",
"type": "boolean"
}
}
},
"recovery.RestoreReceipt": {
"type": "object",
"properties": {
"actor": {
"type": "string"
},
"addonId": {
"type": "string"
},
"at": {
"type": "string"
},
"backupCreated": {
"type": "string"
},
"backupId": {
"type": "string"
},
"lostWrites": {
"description": "LostWrites counts the operations applied after the backup. Their\neffects are gone from the data.",
"type": "integer"
},
"policy": {
"$ref": "#/definitions/recovery.RestoreReport"
},
"previousCopy": {
"description": "PreviousCopy names the directory that holds the data that was replaced.\nIt stays until an administrator purges it.",
"type": "string"
},
"warnings": {
"type": "array",
"items": {
"type": "string"
}
}
}
},
"recovery.RestoreReport": {
"type": "object",
"properties": {
"grantsDropped": {
"type": "integer"
},
"grantsRestored": {
"type": "integer"
},
"historyTruncated": {
"description": "HistoryTruncated says the record of lowered access may have lost\nentries older than the backup, so no grant of the backup was restored.",
"type": "boolean"
},
"proposalsCancelled": {
"description": "ProposalsCancelled and ProposalsUnknown are the proposals of the add-on\nthat no longer can be confirmed, and the ones whose outcome is looked up\nby receipt and never applied again.",
"type": "integer"
},
"proposalsUnknown": {
"type": "integer"
},
"resourcesRestored": {
"type": "integer"
},
"resourcesSkipped": {
"type": "array",
"items": {
"type": "string"
}
},
"revoked": {
"description": "Revoked counts grants dropped or lowered because access was taken away\nafter the backup.",
"type": "integer"
}
}
},
"recovery.RetentionView": {
"type": "object",
"properties": {
"keep": {
"type": "integer"
},
"maxAgeSeconds": {
"type": "integer"
}
}
},
"recovery.Status": {
"type": "object",
"properties": {
"addonId": {
"type": "string"
},
"backups": {
"type": "array",
"items": {
"$ref": "#/definitions/recovery.BackupInfo"
}
},
"lastRestore": {
"$ref": "#/definitions/recovery.RestoreReceipt"
},
"mode": {
"description": "Mode is the ownership mode of the owner: named-user or local-owner.",
"type": "string"
},
"operations": {
"description": "Operations lists the operation IDs whose result is recorded, so a\ncaller that lost an answer can see that its request ran.",
"type": "array",
"items": {
"type": "string"
}
},
"phase": {
"description": "Phase is the phase of an upgrade or restore in progress.",
"type": "string"
},
"previousCopies": {
"description": "PreviousCopies counts the data copies a restore set aside.",
"type": "integer"
},
"reason": {
"type": "string"
},
"recoveryBackup": {
"description": "RecoveryBackup names the copy to restore from.",
"type": "string"
},
"retention": {
"description": "Retention is the policy applied to the backups.",
"allOf": [
{
"$ref": "#/definitions/recovery.RetentionView"
}
]
},
"state": {
"type": "string"
}
}
},
"schema.APIError": {
"type": "object",
"properties": {
"code": {},
"message": {
"type": "string"
},
"param": {
"type": "string"
},
"type": {
"type": "string"
}
}
},
"schema.AnimationInput": {
"type": "object",
"properties": {
"data": {
"type": "string"
},
"type": {
"type": "string"
}
}
},
"schema.AnthropicContentBlock": {
"type": "object",
"properties": {
"content": {},
"id": {
"type": "string"
},
"input": {
"type": "object",
"additionalProperties": {}
},
"is_error": {
"type": "boolean"
},
"name": {
"type": "string"
},
"signature": {
"type": "string"
},
"source": {
"$ref": "#/definitions/schema.AnthropicImageSource"
},
"text": {
"type": "string"
},
"thinking": {
"type": "string"
},
"tool_use_id": {
"type": "string"
},
"type": {
"type": "string"
}
}
},
"schema.AnthropicImageSource": {
"type": "object",
"properties": {
"data": {
"type": "string"
},
"media_type": {
"type": "string"
},
"type": {
"type": "string"
}
}
},
"schema.AnthropicMessage": {
"type": "object",
"properties": {
"content": {},
"role": {
"type": "string"
}
}
},
"schema.AnthropicRequest": {
"type": "object",
"properties": {
"max_tokens": {
"type": "integer"
},
"messages": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.AnthropicMessage"
}
},
"metadata": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"stop_sequences": {
"type": "array",
"items": {
"type": "string"
}
},
"stream": {
"type": "boolean"
},
"system": {
"type": "string"
},
"temperature": {
"type": "number"
},
"thinking": {
"description": "Thinking gates extended-thinking output. We only surface thinking\ncontent blocks when the client explicitly opts in, matching Anthropic's\nAPI where thinking is off unless requested.",
"allOf": [
{
"$ref": "#/definitions/schema.AnthropicThinkingParam"
}
]
},
"tool_choice": {},
"tools": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.AnthropicTool"
}
},
"top_k": {
"type": "integer"
},
"top_p": {
"type": "number"
}
}
},
"schema.AnthropicResponse": {
"type": "object",
"properties": {
"content": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.AnthropicContentBlock"
}
},
"id": {
"type": "string"
},
"model": {
"type": "string"
},
"role": {
"type": "string"
},
"stop_reason": {
"type": "string"
},
"stop_sequence": {
"type": "string"
},
"type": {
"type": "string"
},
"usage": {
"$ref": "#/definitions/schema.AnthropicUsage"
}
}
},
"schema.AnthropicThinkingParam": {
"type": "object",
"properties": {
"budget_tokens": {
"type": "integer"
},
"type": {
"description": "\"enabled\" | \"disabled\"",
"type": "string"
}
}
},
"schema.AnthropicTool": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"input_schema": {
"type": "object",
"additionalProperties": {}
},
"name": {
"type": "string"
}
}
},
"schema.AnthropicUsage": {
"type": "object",
"properties": {
"input_tokens": {
"type": "integer"
},
"output_tokens": {
"type": "integer"
}
}
},
"schema.BackendMonitorRequest": {
"type": "object",
"properties": {
"model": {
"type": "string"
}
}
},
"schema.BackendResponse": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"status_url": {
"type": "string"
}
}
},
"schema.Choice": {
"type": "object",
"properties": {
"delta": {
"$ref": "#/definitions/schema.Message"
},
"finish_reason": {
"type": "string"
},
"index": {
"type": "integer"
},
"logprobs": {
"$ref": "#/definitions/schema.Logprobs"
},
"message": {
"$ref": "#/definitions/schema.Message"
},
"text": {
"type": "string"
}
}
},
"schema.CompressionMetadata": {
"type": "object",
"properties": {
"compressed_tokens": {
"type": "integer"
},
"compressor": {
"type": "string"
},
"dropped_turns": {
"type": "integer"
},
"original_tokens": {
"type": "integer"
},
"overflow_recoveries": {
"type": "integer"
},
"summary_tokens": {
"type": "integer"
}
}
},
"schema.ConvTokenUsage": {
"type": "object",
"properties": {
"completion": {
"type": "integer"
},
"prompt": {
"type": "integer"
},
"total": {
"type": "integer"
}
}
},
"schema.Conversation": {
"type": "object",
"properties": {
"addonToolIds": {
"type": "array",
"items": {
"type": "string"
}
},
"addonTools": {
"type": "boolean"
},
"clientMCPServers": {
"type": "array",
"items": {
"type": "object"
}
},
"contextSize": {
"type": "integer"
},
"createdAt": {
"type": "integer"
},
"history": {
"type": "array",
"items": {
"type": "object"
}
},
"id": {
"type": "string"
},
"localaiAssistant": {
"description": "LocalAIAssistant wires the chat to the in-process admin MCP server\n(the LocalAI Assistant mode of the Chat page).",
"type": "boolean"
},
"mcpMode": {
"type": "boolean"
},
"mcpResources": {
"type": "array",
"items": {
"type": "string"
}
},
"mcpServers": {
"type": "array",
"items": {
"type": "string"
}
},
"metadata": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"name": {
"type": "string"
},
"reasoningEffort": {
"description": "ReasoningEffort is the chat's reasoning_effort choice ('' = model config,\n'none' also hides reasoning traces).",
"type": "string"
},
"systemPrompt": {
"type": "string"
},
"temperature": {
"type": "number"
},
"tokenUsage": {
"$ref": "#/definitions/schema.ConvTokenUsage"
},
"topK": {
"type": "number"
},
"topP": {
"type": "number"
},
"updatedAt": {
"type": "integer"
},
"version": {
"description": "Version is assigned by the server: 1 on the first save, +1 on every\nlater write including the delete. Clients use a non-zero version to\ntell a conversation the server has seen from a never-synced draft.\nIgnored on input.",
"type": "integer"
}
}
},
"schema.ConversationBulkRequest": {
"type": "object",
"properties": {
"conversations": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Conversation"
}
}
}
},
"schema.ConversationBulkResponse": {
"type": "object",
"properties": {
"count": {
"type": "integer"
},
"skipped": {
"type": "array",
"items": {
"type": "string"
}
},
"status": {
"type": "string"
}
}
},
"schema.ConversationList": {
"type": "object",
"properties": {
"conversations": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Conversation"
}
},
"deleted": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ConversationTombstone"
}
}
}
},
"schema.ConversationTombstone": {
"type": "object",
"properties": {
"deletedAt": {
"description": "Unix milliseconds",
"type": "integer"
},
"id": {
"type": "string"
},
"version": {
"type": "integer"
}
}
},
"schema.DecisionAnswer": {
"type": "object",
"properties": {
"choice": {
"description": "Choice preserves the selected string or boolean value. Swagger 2 uses a loose object placeholder for this union.",
"type": "object"
},
"confidence": {
"type": "number"
},
"name": {
"type": "string"
},
"probabilities": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DecisionProbability"
}
},
"probability": {
"type": "number"
},
"score": {
"type": "number"
},
"type": {
"type": "string"
}
}
},
"schema.DecisionChoice": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"value": {
"description": "Value is a string or boolean choice. Swagger 2 uses a loose object placeholder for this union.",
"type": "object"
}
}
},
"schema.DecisionLevel": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"label": {
"type": "string"
}
}
},
"schema.DecisionProbability": {
"type": "object",
"properties": {
"label": {
"type": "string"
},
"probability": {
"type": "number"
},
"value": {
"description": "Value is a string or boolean for choices, or a zero-based integer for score levels. Swagger 2 uses a loose object placeholder for this union.",
"type": "object"
}
}
},
"schema.DecisionQuestion": {
"type": "object",
"properties": {
"choices": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DecisionChoice"
}
},
"instructions": {
"type": "string"
},
"levels": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DecisionLevel"
}
},
"name": {
"type": "string"
},
"type": {
"type": "string"
}
}
},
"schema.DecisionUsage": {
"type": "object",
"properties": {
"input_tokens": {
"type": "integer"
},
"input_tokens_details": {
"type": "object",
"additionalProperties": {
"type": "integer"
}
},
"output_tokens": {
"type": "integer"
},
"output_tokens_details": {
"type": "object",
"additionalProperties": {
"type": "integer"
}
},
"total_tokens": {
"type": "integer"
}
}
},
"schema.DecisionsRequest": {
"type": "object",
"properties": {
"input": {
"description": "Input is a text string or an array of user messages. Swagger 2 uses a loose object placeholder for this union.",
"type": "object"
},
"model": {
"type": "string"
},
"questions": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DecisionQuestion"
}
},
"safety_identifier": {
"type": "string"
}
}
},
"schema.DecisionsResponse": {
"type": "object",
"properties": {
"answers": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DecisionAnswer"
}
},
"model": {
"type": "string"
},
"usage": {
"$ref": "#/definitions/schema.DecisionUsage"
}
}
},
"schema.DepthRequest": {
"type": "object",
"properties": {
"dst": {
"description": "optional output directory for exports (glb/colmap)",
"type": "string"
},
"exports": {
"description": "requested exports: \"glb\", \"colmap\"",
"type": "array",
"items": {
"type": "string"
}
},
"image": {
"description": "URL or base64-encoded image to analyze",
"type": "string"
},
"include_confidence": {
"description": "return the per-pixel confidence map (DualDPT)",
"type": "boolean"
},
"include_depth": {
"description": "return the per-pixel depth map",
"type": "boolean"
},
"include_points": {
"description": "back-project to a 3D point cloud (DualDPT)",
"type": "boolean"
},
"include_pose": {
"description": "return camera extrinsics/intrinsics (DualDPT)",
"type": "boolean"
},
"include_sky": {
"description": "return the per-pixel sky map (mono models)",
"type": "boolean"
},
"model": {
"type": "string"
},
"points_conf_thresh": {
"description": "keep points with confidence \u003e= this threshold",
"type": "number"
}
}
},
"schema.DepthResponse": {
"type": "object",
"properties": {
"confidence": {
"description": "width*height row-major confidence (DualDPT)",
"type": "array",
"items": {
"type": "number"
}
},
"depth": {
"description": "width*height row-major metric depth",
"type": "array",
"items": {
"type": "number"
}
},
"export_paths": {
"description": "paths written for the requested exports",
"type": "array",
"items": {
"type": "string"
}
},
"extrinsics": {
"description": "12 floats, 3x4 row-major (world-to-camera)",
"type": "array",
"items": {
"type": "number"
}
},
"height": {
"type": "integer"
},
"intrinsics": {
"description": "9 floats, 3x3 row-major",
"type": "array",
"items": {
"type": "number"
}
},
"is_metric": {
"description": "depth is in metric units",
"type": "boolean"
},
"num_points": {
"description": "number of 3D points",
"type": "integer"
},
"point_colors": {
"description": "base64-encoded num_points*3 uint8 rgb",
"type": "string"
},
"points": {
"description": "num_points*3 xyz, world space",
"type": "array",
"items": {
"type": "number"
}
},
"sky": {
"description": "width*height row-major sky map (mono)",
"type": "array",
"items": {
"type": "number"
}
},
"width": {
"type": "integer"
}
}
},
"schema.Detection": {
"type": "object",
"properties": {
"class_name": {
"type": "string"
},
"confidence": {
"type": "number"
},
"height": {
"type": "number"
},
"mask": {
"description": "base64-encoded PNG segmentation mask",
"type": "string"
},
"width": {
"type": "number"
},
"x": {
"type": "number"
},
"y": {
"type": "number"
}
}
},
"schema.DetectionRequest": {
"type": "object",
"properties": {
"boxes": {
"description": "Box coordinates as [x1,y1,x2,y2,...] quads",
"type": "array",
"items": {
"type": "number"
}
},
"image": {
"description": "URL or base64-encoded image to analyze",
"type": "string"
},
"model": {
"type": "string"
},
"points": {
"description": "Point coordinates as [x,y,label,...] triples (label: 1=pos, 0=neg)",
"type": "array",
"items": {
"type": "number"
}
},
"prompt": {
"description": "Text prompt (for SAM 3 PCS mode)",
"type": "string"
},
"threshold": {
"description": "Detection confidence threshold",
"type": "number"
}
}
},
"schema.DetectionResponse": {
"type": "object",
"properties": {
"detections": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Detection"
}
}
}
},
"schema.DetokenizeRequest": {
"type": "object",
"properties": {
"model": {
"type": "string"
},
"tokens": {
"description": "token IDs to convert back to text",
"type": "array",
"items": {
"type": "integer"
}
}
}
},
"schema.DetokenizeResponse": {
"type": "object",
"properties": {
"content": {
"description": "detokenized text",
"type": "string"
}
}
},
"schema.DiarizationResult": {
"type": "object",
"properties": {
"duration": {
"type": "number"
},
"language": {
"type": "string"
},
"num_speakers": {
"type": "integer"
},
"segments": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DiarizationSegment"
}
},
"sounds": {
"description": "Sounds is present only when the request set include_sounds. An empty\nlist then means the model ran and heard no event; omitzero keeps a nil\nlist (not requested) out of the payload while an empty one stays.",
"type": "array",
"items": {
"$ref": "#/definitions/schema.DiarizationSound"
}
},
"speaker_profiles": {
"$ref": "#/definitions/schema.SpeakerProfiles"
},
"speakers": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.DiarizationSpeaker"
}
},
"task": {
"type": "string"
}
}
},
"schema.DiarizationSegment": {
"type": "object",
"properties": {
"end": {
"type": "number"
},
"id": {
"type": "integer"
},
"label": {
"type": "string"
},
"name": {
"description": "Name is the registered speaker this segment was matched to, and NameScore\nthe cosine similarity of the match. Both are omitted when the backend did\nnot identify the speaker. Speaker stays the normalized SPEAKER_NN label.",
"type": "string"
},
"name_score": {
"type": "number"
},
"speaker": {
"type": "string"
},
"start": {
"type": "number"
},
"text": {
"type": "string"
}
}
},
"schema.DiarizationSound": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"end": {
"type": "number"
},
"label": {
"type": "string"
},
"start": {
"type": "number"
}
}
},
"schema.DiarizationSpeaker": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"label": {
"type": "string"
},
"name": {
"type": "string"
},
"segment_count": {
"type": "integer"
},
"total_speech_duration": {
"type": "number"
}
}
},
"schema.ElevenLabsSoundGenerationRequest": {
"type": "object",
"properties": {
"bpm": {
"type": "integer"
},
"caption": {
"type": "string"
},
"do_sample": {
"type": "boolean"
},
"duration_seconds": {
"type": "number"
},
"expansion_model": {
"type": "string"
},
"instrumental": {
"description": "Simple mode: use text as description; optional instrumental / vocal_language",
"type": "boolean"
},
"keyscale": {
"type": "string"
},
"language": {
"type": "string"
},
"lyrics": {
"type": "string"
},
"model_id": {
"type": "string"
},
"params": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"prompt_influence": {
"type": "number"
},
"text": {
"type": "string"
},
"think": {
"description": "Advanced mode",
"type": "boolean"
},
"timesignature": {
"type": "string"
},
"vocal_language": {
"type": "string"
}
}
},
"schema.ErrorResponse": {
"type": "object",
"properties": {
"error": {
"$ref": "#/definitions/schema.APIError"
}
}
},
"schema.FaceAnalysis": {
"type": "object",
"properties": {
"age": {
"type": "number"
},
"antispoof_score": {
"type": "number"
},
"dominant_emotion": {
"type": "string"
},
"dominant_gender": {
"type": "string"
},
"dominant_race": {
"type": "string"
},
"emotion": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float32"
}
},
"face_confidence": {
"type": "number"
},
"gender": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float32"
}
},
"is_real": {
"description": "Liveness fields — see FaceVerifyResponse for why these are pointers.",
"type": "boolean"
},
"race": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float32"
}
},
"region": {
"$ref": "#/definitions/schema.FacialArea"
}
}
},
"schema.FaceAnalyzeRequest": {
"type": "object",
"properties": {
"actions": {
"description": "subset of {\"age\",\"gender\",\"emotion\",\"race\"}",
"type": "array",
"items": {
"type": "string"
}
},
"anti_spoofing": {
"type": "boolean"
},
"img": {
"type": "string"
},
"model": {
"type": "string"
}
}
},
"schema.FaceAnalyzeResponse": {
"type": "object",
"properties": {
"faces": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.FaceAnalysis"
}
}
}
},
"schema.FaceEmbedRequest": {
"type": "object",
"properties": {
"img": {
"type": "string"
},
"model": {
"type": "string"
}
}
},
"schema.FaceEmbedResponse": {
"type": "object",
"properties": {
"dim": {
"type": "integer"
},
"embedding": {
"type": "array",
"items": {
"type": "number"
}
},
"model": {
"type": "string"
}
}
},
"schema.FaceForgetRequest": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"model": {
"type": "string"
},
"store": {
"type": "string"
}
}
},
"schema.FaceIdentifyMatch": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"distance": {
"type": "number"
},
"id": {
"type": "string"
},
"labels": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"match": {
"description": "true when distance \u003c= threshold",
"type": "boolean"
},
"name": {
"type": "string"
}
}
},
"schema.FaceIdentifyRequest": {
"type": "object",
"properties": {
"img": {
"type": "string"
},
"model": {
"type": "string"
},
"store": {
"type": "string"
},
"threshold": {
"description": "optional cutoff on distance",
"type": "number"
},
"top_k": {
"type": "integer"
}
}
},
"schema.FaceIdentifyResponse": {
"type": "object",
"properties": {
"matches": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.FaceIdentifyMatch"
}
}
}
},
"schema.FaceRegisterRequest": {
"type": "object",
"properties": {
"embedding": {
"type": "array",
"items": {
"type": "number"
}
},
"img": {
"type": "string"
},
"labels": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"name": {
"type": "string"
},
"registered_at": {
"description": "original enrollment time when replaying a saved embedding",
"type": "string"
},
"store": {
"description": "vector store model; empty = local-store default",
"type": "string"
}
}
},
"schema.FaceRegisterResponse": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"registered_at": {
"type": "string"
}
}
},
"schema.FaceVerifyRequest": {
"type": "object",
"properties": {
"anti_spoofing": {
"type": "boolean"
},
"img1": {
"type": "string"
},
"img2": {
"type": "string"
},
"model": {
"type": "string"
},
"threshold": {
"type": "number"
}
}
},
"schema.FaceVerifyResponse": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"distance": {
"type": "number"
},
"img1_antispoof_score": {
"type": "number"
},
"img1_area": {
"$ref": "#/definitions/schema.FacialArea"
},
"img1_is_real": {
"description": "Liveness fields are only populated when the request set\nanti_spoofing=true. Pointers keep them fully absent from the\nJSON response otherwise, so callers can tell \"not checked\"\napart from \"checked and fake\" (which would collapse to zero\nvalues with plain bool+omitempty).",
"type": "boolean"
},
"img2_antispoof_score": {
"type": "number"
},
"img2_area": {
"$ref": "#/definitions/schema.FacialArea"
},
"img2_is_real": {
"type": "boolean"
},
"model": {
"type": "string"
},
"processing_time_ms": {
"type": "number"
},
"threshold": {
"type": "number"
},
"verified": {
"type": "boolean"
}
}
},
"schema.FacialArea": {
"type": "object",
"properties": {
"h": {
"type": "number"
},
"w": {
"type": "number"
},
"x": {
"type": "number"
},
"y": {
"type": "number"
}
}
},
"schema.ForecastRequest": {
"type": "object",
"properties": {
"clip_nonnegative": {
"description": "null = model default",
"type": "boolean"
},
"max_context": {
"description": "cap on history used; 0 = model default",
"type": "integer"
},
"model": {
"type": "string"
},
"prediction_length": {
"description": "steps to forecast",
"type": "integer"
},
"quantiles": {
"description": "levels to return; default 0.1, 0.5, 0.9",
"type": "array",
"items": {
"type": "number"
}
},
"series": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ForecastSeries"
}
},
"symmetric_averaging": {
"description": "null = model default",
"type": "boolean"
}
}
},
"schema.ForecastResponse": {
"type": "object",
"properties": {
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ForecastSeriesResult"
}
},
"model": {
"type": "string"
},
"object": {
"type": "string"
},
"prediction_length": {
"type": "integer"
},
"quantile_levels": {
"type": "array",
"items": {
"type": "number"
}
}
}
},
"schema.ForecastSeries": {
"type": "object",
"properties": {
"id": {
"description": "echoed back in the matching result",
"type": "string"
},
"known_categorical": {
"description": "KnownCategorical is [C][T+H]: integer category codes for the future too.",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float64"
}
}
},
"known_covariates": {
"description": "KnownCovariates is [C][T+H]: numeric covariates known for the future too.",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float64"
}
}
},
"past_categorical": {
"description": "PastCategorical is [C][T]: integer category codes for the past.",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float64"
}
}
},
"past_covariates": {
"description": "PastCovariates is [C][T]: numeric covariates known only for the past.",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float64"
}
}
},
"target": {
"description": "[V][T] or [T], null = missing",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float64"
}
}
}
}
},
"schema.ForecastSeriesResult": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"index": {
"description": "position of the series in the request",
"type": "integer"
},
"variates": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ForecastVariate"
}
}
}
},
"schema.ForecastVariate": {
"type": "object",
"properties": {
"index": {
"type": "integer"
},
"values": {
"description": "Values is [Q][H]: Values[q][h] is quantile_levels[q] at step h.",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "number",
"format": "float32"
}
}
}
}
},
"schema.FunctionCall": {
"type": "object",
"properties": {
"arguments": {
"type": "string"
},
"name": {
"type": "string"
}
}
},
"schema.GalleryResponse": {
"type": "object",
"properties": {
"estimated_size_bytes": {
"type": "integer"
},
"estimated_size_display": {
"type": "string"
},
"estimated_vram_bytes": {
"type": "integer"
},
"estimated_vram_display": {
"type": "string"
},
"status": {
"type": "string"
},
"uuid": {
"type": "string"
}
}
},
"schema.InputTokensDetails": {
"type": "object",
"properties": {
"image_tokens": {
"type": "integer"
},
"text_tokens": {
"type": "integer"
}
}
},
"schema.Item": {
"type": "object",
"properties": {
"b64_json": {
"type": "string"
},
"index": {
"type": "integer"
},
"object": {
"type": "string"
},
"url": {
"description": "Images",
"type": "string"
}
}
},
"schema.JINADocumentResult": {
"type": "object",
"properties": {
"document": {
"$ref": "#/definitions/schema.JINAText"
},
"index": {
"type": "integer"
},
"relevance_score": {
"type": "number"
}
}
},
"schema.JINARerankRequest": {
"type": "object",
"properties": {
"backend": {
"type": "string"
},
"documents": {
"type": "array",
"items": {
"type": "string"
}
},
"model": {
"type": "string"
},
"query": {
"type": "string"
},
"top_n": {
"type": "integer"
}
}
},
"schema.JINARerankResponse": {
"type": "object",
"properties": {
"model": {
"type": "string"
},
"results": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.JINADocumentResult"
}
},
"usage": {
"$ref": "#/definitions/schema.JINAUsageInfo"
}
}
},
"schema.JINAText": {
"type": "object",
"properties": {
"text": {
"type": "string"
}
}
},
"schema.JINAUsageInfo": {
"type": "object",
"properties": {
"prompt_tokens": {
"type": "integer"
},
"total_tokens": {
"type": "integer"
}
}
},
"schema.Job": {
"type": "object",
"properties": {
"audios": {
"description": "List of audio URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
},
"completed_at": {
"type": "string"
},
"created_at": {
"type": "string"
},
"error": {
"description": "Error message if failed",
"type": "string"
},
"files": {
"description": "List of file URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
},
"id": {
"description": "UUID",
"type": "string"
},
"images": {
"description": "Multimedia content (for manual execution)\nCan contain URLs or base64-encoded data URIs",
"type": "array",
"items": {
"type": "string"
}
},
"parameters": {
"description": "Template parameters",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"result": {
"description": "Agent response",
"type": "string"
},
"started_at": {
"type": "string"
},
"status": {
"description": "pending, running, completed, failed, cancelled",
"allOf": [
{
"$ref": "#/definitions/schema.JobStatus"
}
]
},
"task_id": {
"description": "Reference to Task",
"type": "string"
},
"traces": {
"description": "Execution traces (reasoning, tool calls, tool results)",
"type": "array",
"items": {
"$ref": "#/definitions/schema.JobTrace"
}
},
"triggered_by": {
"description": "\"manual\", \"cron\", \"api\"",
"type": "string"
},
"videos": {
"description": "List of video URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
},
"webhook_error": {
"description": "Error if webhook failed",
"type": "string"
},
"webhook_sent": {
"description": "Webhook delivery tracking",
"type": "boolean"
},
"webhook_sent_at": {
"type": "string"
}
}
},
"schema.JobExecutionRequest": {
"type": "object",
"properties": {
"audios": {
"description": "List of audio URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
},
"files": {
"description": "List of file URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
},
"images": {
"description": "Multimedia content (optional, for manual execution)\nCan contain URLs or base64-encoded data URIs",
"type": "array",
"items": {
"type": "string"
}
},
"parameters": {
"description": "Optional, for templating",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"task_id": {
"description": "Required",
"type": "string"
},
"videos": {
"description": "List of video URLs or base64 strings",
"type": "array",
"items": {
"type": "string"
}
}
}
},
"schema.JobExecutionResponse": {
"type": "object",
"properties": {
"job_id": {
"description": "unique job identifier",
"type": "string"
},
"status": {
"description": "initial status (pending)",
"type": "string"
},
"url": {
"description": "URL to poll for job status",
"type": "string"
}
}
},
"schema.JobStatus": {
"type": "string",
"enum": [
"pending",
"running",
"completed",
"failed",
"cancelled"
],
"x-enum-varnames": [
"JobStatusPending",
"JobStatusRunning",
"JobStatusCompleted",
"JobStatusFailed",
"JobStatusCancelled"
]
},
"schema.JobTrace": {
"type": "object",
"properties": {
"arguments": {
"description": "Tool arguments or result data",
"type": "object",
"additionalProperties": {}
},
"content": {
"description": "The actual trace content",
"type": "string"
},
"timestamp": {
"description": "When this trace occurred",
"type": "string"
},
"tool_name": {
"description": "Tool name (for tool_call/tool_result)",
"type": "string"
},
"type": {
"description": "\"reasoning\", \"tool_call\", \"tool_result\", \"status\"",
"type": "string"
}
}
},
"schema.KnownBackend": {
"type": "object",
"properties": {
"auto_detect": {
"type": "boolean"
},
"description": {
"type": "string"
},
"installed": {
"description": "Installed is true when the backend is currently present on disk — i.e. it\nappears in gallery.ListSystemBackends(systemState). Importer-registered or\ncurated pref-only backends default to false unless they also show up on\ndisk. The import form uses this to warn users that submitting an import\nmay trigger an automatic backend download.",
"type": "boolean"
},
"modality": {
"type": "string"
},
"name": {
"type": "string"
}
}
},
"schema.LogprobContent": {
"type": "object",
"properties": {
"bytes": {
"type": "array",
"items": {
"type": "integer"
}
},
"id": {
"type": "integer"
},
"logprob": {
"type": "number"
},
"token": {
"type": "string"
},
"top_logprobs": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.LogprobContent"
}
}
}
},
"schema.Logprobs": {
"type": "object",
"properties": {
"content": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.LogprobContent"
}
}
}
},
"schema.LogprobsValue": {
"type": "object",
"properties": {
"enabled": {
"description": "true if logprobs should be returned",
"type": "boolean"
}
}
},
"schema.Message": {
"type": "object",
"properties": {
"content": {
"description": "The message content"
},
"function_call": {
"description": "A result of a function call"
},
"name": {
"description": "The message name (used for tools calls)",
"type": "string"
},
"reasoning": {
"description": "Reasoning content extracted from \u003cthinking\u003e...\u003c/thinking\u003e tags",
"type": "string"
},
"role": {
"description": "The message role",
"type": "string"
},
"tool_call_id": {
"type": "string"
},
"tool_calls": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ToolCall"
}
}
}
},
"schema.Model3DAnimationRequest": {
"type": "object",
"properties": {
"inputs": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/schema.AnimationInput"
}
},
"model": {
"type": "string"
},
"params": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"response_format": {
"type": "string"
}
}
},
"schema.Model3DRequest": {
"description": "3D asset generation request body. Generation is image-conditioned",
"type": "object",
"properties": {
"background": {
"description": "background handling: auto|keep|black|white",
"type": "string"
},
"cfg_scale": {
"description": "classifier-free guidance scale (backend default 7.5)",
"type": "number"
},
"image": {
"description": "conditioning image: URL, base64, or data URI (required)",
"type": "string"
},
"model": {
"type": "string"
},
"params": {
"description": "backend-specific generation parameters",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"quality": {
"description": "mesh pipeline: auto|coarse|512|1024",
"type": "string"
},
"response_format": {
"description": "output format (url or b64_json)",
"type": "string"
},
"seed": {
"description": "random seed; \u003c=0 picks a random seed",
"type": "integer"
},
"step": {
"description": "flow sampling steps (backend default 12)",
"type": "integer"
},
"texture_steps": {
"description": "texture flow sampling steps (backend default 12)",
"type": "integer"
}
}
},
"schema.ModelCapabilities": {
"type": "object",
"properties": {
"capabilities": {
"description": "Capabilities are canonical usecase strings (e.g. chat, vision, transcript,\ntts, embeddings, image, video) plus the modifiers \"tools\" and \"thinking\".",
"type": "array",
"items": {
"type": "string"
}
},
"context_size": {
"description": "ContextSize is the effective context window in tokens the backend will\nrun with: the configured context_size, or the default when unset. 0 means\nthe value is unknown (e.g. a loose file with no config). Clients can use\nthis to size their context budget for auto-compaction and pruning.",
"type": "integer"
},
"id": {
"type": "string"
},
"input_modalities": {
"description": "InputModalities is the subset of {text,image,audio,video} the model accepts.",
"type": "array",
"items": {
"type": "string"
}
},
"object": {
"type": "string"
},
"output_modalities": {
"description": "OutputModalities is the subset of {text,image,audio,video} the model produces.",
"type": "array",
"items": {
"type": "string"
}
},
"three_d_operations": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ThreeDOperation"
}
},
"upscale_scale": {
"type": "integer"
}
}
},
"schema.ModelCapabilitiesResponse": {
"type": "object",
"properties": {
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ModelCapabilities"
}
},
"object": {
"type": "string"
}
}
},
"schema.ModelLoadCancelRequest": {
"type": "object",
"properties": {
"job_id": {
"type": "string"
}
}
},
"schema.ModelLoadCancelResponse": {
"type": "object",
"properties": {
"current_job_id": {
"type": "string"
},
"job_id": {
"type": "string"
},
"model": {
"type": "string"
},
"retry_after": {
"type": "integer"
},
"state": {
"type": "string"
}
}
},
"schema.ModelLoadRequest": {
"type": "object",
"properties": {
"model": {
"type": "string"
}
}
},
"schema.ModelLoadResponse": {
"type": "object",
"properties": {
"loaded": {
"description": "Loaded lists the model names actually resident in memory after the call.\nFor a pipeline model these are its sub-models, not the pipeline name.",
"type": "array",
"items": {
"type": "string"
}
},
"message": {
"description": "Message is a short human-readable status (\"model loaded\", or an error).",
"type": "string"
}
}
},
"schema.ModelLoadingStatus": {
"type": "object",
"properties": {
"bytes_sent": {
"type": "integer"
},
"cancel_requested": {
"description": "CancelRequested is true when an administrator cancelled the attempt.",
"type": "boolean"
},
"eta_seconds": {
"description": "ETASeconds is omitted rather than guessed until enough bytes have moved\nfor the observed rate to mean anything. A confidently wrong ETA on a\ntwenty-minute wait is worse than none.",
"type": "integer"
},
"file_index": {
"type": "integer"
},
"job_id": {
"description": "JobID names the load attempt. A cancel must quote it: it is the\nprecondition that keeps a cancel from hitting a replacement attempt.",
"type": "string"
},
"last_error": {
"description": "LastError is the cause of a failed attempt.",
"type": "string"
},
"lease_expires_in": {
"description": "LeaseExpiresIn is the seconds left on the owner's lease. It is negative\nwhen the lease already ran out, which means the owner is gone.",
"type": "integer"
},
"model": {
"type": "string"
},
"node": {
"type": "string"
},
"progress": {
"type": "number"
},
"retry_after": {
"description": "RetryAfter is the seconds until a new load may start, for a failed attempt.",
"type": "integer"
},
"state": {
"type": "string"
},
"stop_deadline": {
"type": "string"
},
"stopping": {
"description": "Stopping is true while the remote work of a failed attempt is not yet\nconfirmed ended. StopDeadline is when the model is released regardless.",
"type": "boolean"
},
"total_bytes": {
"type": "integer"
},
"total_files": {
"type": "integer"
}
}
},
"schema.ModelsDataResponse": {
"type": "object",
"properties": {
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.OpenAIModel"
}
},
"object": {
"type": "string"
}
}
},
"schema.ModerationRequest": {
"type": "object",
"properties": {
"input": {
"type": "array",
"items": {
"type": "string"
}
},
"model": {
"type": "string"
}
}
},
"schema.ModerationResponse": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"model": {
"type": "string"
},
"results": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ModerationResult"
}
}
}
},
"schema.ModerationResult": {
"type": "object",
"properties": {
"categories": {
"type": "object",
"additionalProperties": {
"type": "boolean"
}
},
"category_applied_input_types": {
"type": "object",
"additionalProperties": {
"type": "array",
"items": {
"type": "string"
}
}
},
"category_scores": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
},
"flagged": {
"type": "boolean"
}
}
},
"schema.MultimediaSourceConfig": {
"type": "object",
"properties": {
"headers": {
"description": "Custom headers for HTTP request (e.g., Authorization)",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"type": {
"description": "\"image\", \"video\", \"audio\", \"file\"",
"type": "string"
},
"url": {
"description": "URL to fetch from",
"type": "string"
}
}
},
"schema.NodeData": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"lastSeen": {
"type": "string"
},
"name": {
"type": "string"
},
"serviceID": {
"type": "string"
},
"tunnelAddress": {
"type": "string"
}
}
},
"schema.ORAnnotation": {
"type": "object",
"properties": {
"end_index": {
"type": "integer"
},
"start_index": {
"type": "integer"
},
"title": {
"type": "string"
},
"type": {
"description": "url_citation",
"type": "string"
},
"url": {
"type": "string"
}
}
},
"schema.ORContentPart": {
"type": "object",
"properties": {
"annotations": {
"description": "REQUIRED for output_text - must always be present (use [])",
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORAnnotation"
}
},
"detail": {
"description": "low|high|auto for images",
"type": "string"
},
"file_data": {
"type": "string"
},
"file_url": {
"type": "string"
},
"filename": {
"type": "string"
},
"image_url": {
"type": "string"
},
"logprobs": {
"description": "REQUIRED for output_text - must always be present (use [])",
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORLogProb"
}
},
"refusal": {
"type": "string"
},
"text": {
"description": "REQUIRED for output_text - must always be present (even if empty)",
"type": "string"
},
"type": {
"description": "input_text|input_image|input_file|output_text|refusal",
"type": "string"
}
}
},
"schema.ORError": {
"type": "object",
"properties": {
"code": {
"type": "string"
},
"message": {
"type": "string"
},
"param": {
"type": "string"
},
"type": {
"description": "invalid_request|not_found|server_error|model_error|too_many_requests",
"type": "string"
}
}
},
"schema.ORFunctionTool": {
"type": "object",
"properties": {
"description": {
"type": "string"
},
"name": {
"type": "string"
},
"parameters": {
"type": "object",
"additionalProperties": {}
},
"strict": {
"description": "Always include in response",
"type": "boolean"
},
"type": {
"type": "string"
}
}
},
"schema.ORIncompleteDetails": {
"type": "object",
"properties": {
"reason": {
"type": "string"
}
}
},
"schema.ORInputTokensDetails": {
"type": "object",
"properties": {
"cached_tokens": {
"description": "Always include, even if 0",
"type": "integer"
}
}
},
"schema.ORItemField": {
"type": "object",
"properties": {
"arguments": {
"type": "string"
},
"call_id": {
"description": "Function call fields",
"type": "string"
},
"content": {
"description": "string or []ORContentPart for messages"
},
"encrypted_content": {
"description": "Provider-specific encrypted content",
"type": "string"
},
"id": {
"description": "Present for all output items",
"type": "string"
},
"name": {
"type": "string"
},
"output": {
"description": "Function call output fields"
},
"role": {
"description": "Message fields",
"type": "string"
},
"status": {
"description": "in_progress|completed|incomplete",
"type": "string"
},
"summary": {
"description": "Reasoning fields (for type == \"reasoning\")",
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORContentPart"
}
},
"type": {
"description": "message|function_call|function_call_output|reasoning|item_reference",
"type": "string"
}
}
},
"schema.ORLogProb": {
"type": "object",
"properties": {
"bytes": {
"type": "array",
"items": {
"type": "integer"
}
},
"logprob": {
"type": "number"
},
"token": {
"type": "string"
},
"top_logprobs": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORTopLogProb"
}
}
}
},
"schema.OROutputTokensDetails": {
"type": "object",
"properties": {
"reasoning_tokens": {
"description": "Always include, even if 0",
"type": "integer"
}
}
},
"schema.ORReasoning": {
"type": "object",
"properties": {
"effort": {
"type": "string"
},
"summary": {
"type": "string"
}
}
},
"schema.ORReasoningParam": {
"type": "object",
"properties": {
"effort": {
"description": "\"none\"|\"low\"|\"medium\"|\"high\"|\"xhigh\"",
"type": "string"
},
"summary": {
"description": "\"auto\"|\"concise\"|\"detailed\"",
"type": "string"
}
}
},
"schema.ORResponseResource": {
"type": "object",
"properties": {
"background": {
"type": "boolean"
},
"completed_at": {
"description": "Required: present as number or null",
"type": "integer"
},
"created_at": {
"type": "integer"
},
"error": {
"description": "Always present, null if no error",
"allOf": [
{
"$ref": "#/definitions/schema.ORError"
}
]
},
"frequency_penalty": {
"type": "number"
},
"id": {
"type": "string"
},
"incomplete_details": {
"description": "Always present, null if complete",
"allOf": [
{
"$ref": "#/definitions/schema.ORIncompleteDetails"
}
]
},
"instructions": {
"type": "string"
},
"max_output_tokens": {
"type": "integer"
},
"max_tool_calls": {
"description": "nullable",
"type": "integer"
},
"metadata": {
"description": "Metadata and operational flags",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"object": {
"description": "always \"response\"",
"type": "string"
},
"output": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORItemField"
}
},
"parallel_tool_calls": {
"type": "boolean"
},
"presence_penalty": {
"type": "number"
},
"previous_response_id": {
"type": "string"
},
"prompt_cache_key": {
"description": "nullable",
"type": "string"
},
"reasoning": {
"description": "nullable",
"allOf": [
{
"$ref": "#/definitions/schema.ORReasoning"
}
]
},
"safety_identifier": {
"description": "Safety and caching",
"type": "string"
},
"service_tier": {
"type": "string"
},
"status": {
"description": "in_progress|completed|failed|incomplete",
"type": "string"
},
"store": {
"type": "boolean"
},
"temperature": {
"description": "Sampling parameters (always required)",
"type": "number"
},
"text": {
"description": "Text format configuration",
"allOf": [
{
"$ref": "#/definitions/schema.ORTextConfig"
}
]
},
"tool_choice": {},
"tools": {
"description": "Tool-related fields",
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORFunctionTool"
}
},
"top_logprobs": {
"description": "Default to 0",
"type": "integer"
},
"top_p": {
"type": "number"
},
"truncation": {
"description": "Truncation and reasoning",
"type": "string"
},
"usage": {
"description": "Usage statistics",
"allOf": [
{
"$ref": "#/definitions/schema.ORUsage"
}
]
}
}
},
"schema.ORTextConfig": {
"type": "object",
"properties": {
"format": {
"$ref": "#/definitions/schema.ORTextFormat"
}
}
},
"schema.ORTextFormat": {
"type": "object",
"properties": {
"type": {
"description": "\"text\" or \"json_schema\"",
"type": "string"
}
}
},
"schema.ORTopLogProb": {
"type": "object",
"properties": {
"bytes": {
"type": "array",
"items": {
"type": "integer"
}
},
"logprob": {
"type": "number"
},
"token": {
"type": "string"
}
}
},
"schema.ORUsage": {
"type": "object",
"properties": {
"input_tokens": {
"type": "integer"
},
"input_tokens_details": {
"description": "Always present",
"allOf": [
{
"$ref": "#/definitions/schema.ORInputTokensDetails"
}
]
},
"output_tokens": {
"type": "integer"
},
"output_tokens_details": {
"description": "Always present",
"allOf": [
{
"$ref": "#/definitions/schema.OROutputTokensDetails"
}
]
},
"total_tokens": {
"type": "integer"
}
}
},
"schema.OpenAIModel": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"object": {
"type": "string"
}
}
},
"schema.OpenAIRequest": {
"type": "object",
"required": [
"file"
],
"properties": {
"backend": {
"type": "string"
},
"batch": {
"description": "Custom parameters - not present in the OpenAI API",
"type": "integer"
},
"clip_skip": {
"description": "Diffusers",
"type": "integer"
},
"echo": {
"type": "boolean"
},
"encoding_format": {
"description": "Embedding encoding format: \"float\" (default) or \"base64\" (OpenAI Node.js SDK default)",
"type": "string"
},
"file": {
"description": "whisper",
"type": "string"
},
"files": {
"description": "Multiple input images for img2img or inpainting",
"type": "array",
"items": {
"type": "string"
}
},
"frequency_penalty": {
"type": "number"
},
"function_call": {
"description": "might be a string or an object"
},
"functions": {
"description": "A list of available functions to call",
"type": "array",
"items": {
"$ref": "#/definitions/functions.Function"
}
},
"grammar": {
"description": "A grammar to constrain the LLM output",
"type": "string"
},
"grammar_json_functions": {
"$ref": "#/definitions/functions.JSONFunctionStructure"
},
"ignore_eos": {
"type": "boolean"
},
"include_sounds": {
"type": "boolean"
},
"include_speaker_profiles": {
"type": "boolean"
},
"include_text": {
"type": "boolean"
},
"input": {},
"instruction": {
"description": "Edit endpoint",
"type": "string"
},
"language": {
"description": "Also part of the OpenAI official spec",
"type": "string"
},
"logit_bias": {
"description": "Map of token IDs to bias values (-100 to 100)",
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
},
"logprobs": {
"description": "OpenAI API logprobs parameters\nlogprobs: boolean - if true, returns log probabilities of each output token\ntop_logprobs: integer 0-20 - number of most likely tokens to return at each token position",
"allOf": [
{
"$ref": "#/definitions/schema.LogprobsValue"
}
]
},
"max_completion_tokens": {
"description": "MaxCompletionTokens is the modern alias for max_tokens\n(OpenAI deprecated max_tokens; gpt-5 / o-series reject it).\nAccepted on the wire so up-to-date clients can use the new\nname; the request middleware collapses it into Maxtokens so\ninternal code reads exactly one field.",
"type": "integer"
},
"max_tokens": {
"type": "integer"
},
"messages": {
"description": "Messages is read only by chat/completion API calls",
"type": "array",
"items": {
"$ref": "#/definitions/schema.Message"
}
},
"metadata": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"min_p": {
"type": "number"
},
"model": {
"type": "string"
},
"model_base_name": {
"type": "string"
},
"n": {
"description": "Also part of the OpenAI official spec. use it for returning multiple results",
"type": "integer"
},
"n_keep": {
"type": "integer"
},
"negative_prompt": {
"description": "NegativePrompt for image generation (matches Stable Diffusion WebUI /\nvLLM-Omni conventions). Combined, comma-separated, with any \"|\"-suffixed\nnegative tags in Prompt.",
"type": "string"
},
"negative_prompt_scale": {
"type": "number"
},
"pooling": {
"description": "Pooling is a LocalAI extension for /v1/embeddings: how the backend's\nper-token vectors are reduced to a single embedding. \"\" or \"backend\"\nleaves pooling to the inference backend (the pre-existing behavior);\n\"mean\", \"last\" and \"decayed_mean\" pool Go-side from raw per-token\nvectors (the backend must run with the \"pooling:none\" option, which\nmodel configs get automatically when this is set).",
"type": "string"
},
"pooling_half_life_tokens": {
"description": "PoolingHalfLifeTokens is a LocalAI extension for /v1/embeddings: the\nhalf-life (in tokens) of the \"decayed_mean\" pooling scheme — a token's\nweight halves every this-many positions counting back from the end of\nthe conversation. Defaults to 256 when unset.",
"type": "integer"
},
"presence_penalty": {
"type": "number"
},
"prompt": {
"description": "Prompt is read only by completion/image API calls"
},
"quality": {
"description": "Image (not supported by OpenAI)",
"type": "string"
},
"reasoning_effort": {
"type": "string"
},
"ref_images": {
"description": "Reference images for models that support them (e.g., Flux Kontext)",
"type": "array",
"items": {
"type": "string"
}
},
"repeat_last_n": {
"type": "integer"
},
"repeat_penalty": {
"type": "number"
},
"response_format": {
"description": "whisper/image"
},
"rope_freq_base": {
"type": "number"
},
"rope_freq_scale": {
"type": "number"
},
"seed": {
"type": "integer"
},
"size": {
"description": "image",
"type": "string"
},
"step": {
"type": "integer"
},
"stop": {},
"stream": {
"type": "boolean"
},
"stream_options": {
"description": "StreamOptions opts into OpenAI streaming extensions, e.g. include_usage.",
"allOf": [
{
"$ref": "#/definitions/schema.StreamOptions"
}
]
},
"temperature": {
"type": "number"
},
"tfz": {
"type": "number"
},
"tokenizer": {
"description": "RWKV (?)",
"type": "string"
},
"tool_choice": {},
"tools": {
"type": "array",
"items": {
"$ref": "#/definitions/functions.Tool"
}
},
"top_k": {
"type": "integer"
},
"top_logprobs": {
"description": "Number of top logprobs per token (0-20)",
"type": "integer"
},
"top_p": {
"description": "Common options between all the API calls, part of the OpenAI spec",
"type": "number"
},
"translate": {
"description": "Only for audio transcription",
"type": "boolean"
},
"typical_p": {
"type": "number"
}
}
},
"schema.OpenAIResponse": {
"type": "object",
"properties": {
"choices": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Choice"
}
},
"created": {
"type": "integer"
},
"data": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.Item"
}
},
"id": {
"type": "string"
},
"metadata": {
"type": "object"
},
"model": {
"type": "string"
},
"object": {
"type": "string"
},
"usage": {
"description": "Usage is intentionally a pointer with omitempty: per the OpenAI\nchat-completion streaming spec, intermediate chunks must not carry\na ` + "`" + `usage` + "`" + ` field. Marshalling a value-typed usage would emit\n` + "`" + `\"usage\":{\"prompt_tokens\":0,...}` + "`" + ` on every chunk and break\nOpenAI-SDK consumers that filter on a truthy ` + "`" + `result.usage` + "`" + `\n(continuedev/continue, Kilo Code, Roo Code, etc.).",
"allOf": [
{
"$ref": "#/definitions/schema.OpenAIUsage"
}
]
}
}
},
"schema.OpenAIUsage": {
"type": "object",
"properties": {
"completion_tokens": {
"type": "integer"
},
"compression_meta": {
"$ref": "#/definitions/schema.CompressionMetadata"
},
"input_tokens": {
"description": "Fields for image generation API compatibility",
"type": "integer"
},
"input_tokens_details": {
"$ref": "#/definitions/schema.InputTokensDetails"
},
"output_tokens": {
"type": "integer"
},
"prompt_tokens": {
"type": "integer"
},
"timing_prompt_processing": {
"description": "Extra timing data, disabled by default as is't not a part of OpenAI specification",
"type": "number"
},
"timing_token_generation": {
"type": "number"
},
"total_tokens": {
"type": "integer"
}
}
},
"schema.OpenResponsesRequest": {
"type": "object",
"properties": {
"allowed_tools": {
"description": "Restrict which tools can be invoked",
"type": "array",
"items": {
"type": "string"
}
},
"background": {
"description": "Run request in background",
"type": "boolean"
},
"frequency_penalty": {
"description": "Frequency penalty (-2.0 to 2.0)",
"type": "number"
},
"include": {
"description": "What to include in response",
"type": "array",
"items": {
"type": "string"
}
},
"input": {
"description": "string or []ORItemParam"
},
"instructions": {
"type": "string"
},
"logit_bias": {
"description": "OpenAI-compatible extensions (not in Open Responses spec)",
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
},
"max_output_tokens": {
"type": "integer"
},
"max_tool_calls": {
"description": "Maximum number of tool calls",
"type": "integer"
},
"metadata": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"parallel_tool_calls": {
"description": "Allow parallel tool calls",
"type": "boolean"
},
"presence_penalty": {
"description": "Presence penalty (-2.0 to 2.0)",
"type": "number"
},
"previous_response_id": {
"type": "string"
},
"reasoning": {
"$ref": "#/definitions/schema.ORReasoningParam"
},
"service_tier": {
"description": "\"auto\"|\"default\"|priority hint",
"type": "string"
},
"store": {
"description": "Whether to store the response",
"type": "boolean"
},
"stream": {
"type": "boolean"
},
"temperature": {
"type": "number"
},
"text_format": {
"description": "Additional parameters from spec"
},
"tool_choice": {
"description": "\"auto\"|\"required\"|\"none\"|{type:\"function\",name:\"...\"}"
},
"tools": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ORFunctionTool"
}
},
"top_logprobs": {
"description": "Number of top logprobs to return",
"type": "integer"
},
"top_p": {
"type": "number"
},
"truncation": {
"description": "\"auto\"|\"disabled\"",
"type": "string"
}
}
},
"schema.P2PNodesResponse": {
"type": "object",
"properties": {
"federated_nodes": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.NodeData"
}
},
"llama_cpp_nodes": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.NodeData"
}
},
"mlx_nodes": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.NodeData"
}
}
}
},
"schema.PIIAnalyzeRequest": {
"type": "object",
"properties": {
"detectors": {
"description": "Detectors names the detector models to run (NER and/or pattern). Takes\nprecedence over Model.",
"type": "array",
"items": {
"type": "string"
}
},
"model": {
"description": "Model is a consuming model whose effective PII policy (own\npii.detectors, else the instance default detectors; PII must be\nenabled) is used when Detectors is empty.",
"type": "string"
},
"reveal": {
"description": "Reveal includes the per-entity hash_prefix in the response. Honoured\nonly for admin callers; ignored otherwise. The raw matched value is\nnever returned regardless.",
"type": "boolean"
},
"text": {
"description": "Text is the string to scan. Bounded only by the server's global HTTP\nbody limit.",
"type": "string"
}
}
},
"schema.PIIAnalyzeResponse": {
"type": "object",
"properties": {
"blocked": {
"type": "boolean"
},
"correlation_id": {
"type": "string"
},
"entities": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.PIIEntity"
}
}
}
},
"schema.PIIEntity": {
"type": "object",
"properties": {
"action": {
"type": "string"
},
"end": {
"type": "integer"
},
"entity_type": {
"type": "string"
},
"hash_prefix": {
"type": "string"
},
"score": {
"type": "number"
},
"source": {
"type": "string"
},
"start": {
"type": "integer"
}
}
},
"schema.PIIRedactResponse": {
"type": "object",
"properties": {
"blocked": {
"type": "boolean"
},
"correlation_id": {
"type": "string"
},
"entities": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.PIIEntity"
}
},
"masked": {
"type": "boolean"
},
"redacted_text": {
"type": "string"
}
}
},
"schema.RouterCorpusAddRequest": {
"type": "object",
"properties": {
"entries": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.RouterCorpusEntry"
}
}
}
},
"schema.RouterCorpusAddResponse": {
"type": "object",
"properties": {
"added": {
"description": "Added is how many entries were embedded, persisted, and indexed.",
"type": "integer"
},
"label_counts": {
"description": "LabelCounts is the per-label exemplar count after the call.",
"type": "object",
"additionalProperties": {
"type": "integer"
}
},
"router": {
"type": "string"
},
"skipped": {
"description": "Skipped counts entries whose text was already in the corpus —\nduplicates are rejected rather than double-weighted.",
"type": "integer"
},
"total": {
"description": "Total is the corpus size after the call.",
"type": "integer"
}
}
},
"schema.RouterCorpusClearResponse": {
"type": "object",
"properties": {
"cleared": {
"type": "integer"
},
"router": {
"type": "string"
}
}
},
"schema.RouterCorpusEntry": {
"type": "object",
"properties": {
"labels": {
"type": "array",
"items": {
"type": "string"
}
},
"text": {
"type": "string"
}
}
},
"schema.RouterCorpusStatsResponse": {
"type": "object",
"properties": {
"embedding_model": {
"type": "string"
},
"embedding_models": {
"description": "EmbeddingModels lists the embedder fingerprints present in the\npersisted corpus; more than one means part of the corpus is\npending re-embedding on the next load.",
"type": "array",
"items": {
"type": "string"
}
},
"label_counts": {
"type": "object",
"additionalProperties": {
"type": "integer"
}
},
"router": {
"type": "string"
},
"store_name": {
"type": "string"
},
"total": {
"type": "integer"
}
}
},
"schema.RouterDecideNeighbor": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"labels": {
"type": "array",
"items": {
"type": "string"
}
},
"similarity": {
"type": "number"
}
}
},
"schema.RouterDecideRequest": {
"type": "object",
"properties": {
"input": {
"description": "Input is the user-visible prompt text to classify. Required.\nSchema-shape extraction (chat-message concatenation, etc.) is\nthe caller's responsibility — matches the Probe contract used\nby the in-band middleware.",
"type": "string"
},
"router": {
"description": "Router is the name of the router model (a ModelConfig with a\n` + "`" + `router:` + "`" + ` block). Required.",
"type": "string"
}
}
},
"schema.RouterDecideResponse": {
"type": "object",
"properties": {
"cache_similarity": {
"description": "CacheSimilarity carries the cosine similarity of the cache hit\n(0 when not cached).",
"type": "number"
},
"cached": {
"description": "Cached is true when the decision came from the L2 embedding\ncache rather than a fresh classifier run.",
"type": "boolean"
},
"candidate": {
"description": "Candidate is the model that would be routed to. Empty when no\ncandidate covers Labels AND no fallback is configured.",
"type": "string"
},
"classifier": {
"description": "Classifier is the classifier name that produced the decision\n(e.g. \"score\").",
"type": "string"
},
"fallback": {
"description": "Fallback is true when Candidate is the router's configured\nfallback because no candidate covered Labels. Lets callers\ndistinguish \"matched\" from \"fell back\" without comparing names.",
"type": "boolean"
},
"labels": {
"description": "Labels is the set of active policy labels.",
"type": "array",
"items": {
"type": "string"
}
},
"latency_ms": {
"description": "LatencyMs is the classifier's wall-clock cost.",
"type": "integer"
},
"nearest_similarity": {
"description": "NearestSimilarity is the cosine similarity of the closest KNN\ncorpus entry — populated by the knn classifier even when the\ndecision fell back because the probe was out of corpus range.\n0 for other classifiers.",
"type": "number"
},
"neighbors": {
"description": "Neighbors lists the corpus entries the knn classifier retrieved,\nby descending similarity, including ones below the similarity\ngate. Empty for other classifiers.",
"type": "array",
"items": {
"$ref": "#/definitions/schema.RouterDecideNeighbor"
}
},
"router": {
"description": "Router echoes the requested router model.",
"type": "string"
},
"score": {
"description": "Score is the top label's softmax probability (the\nclassifier-side confidence signal).",
"type": "number"
}
}
},
"schema.SlotCacheRequest": {
"type": "object",
"properties": {
"id": {
"description": "restore only: an ID issued by a save for this model",
"type": "string"
},
"model": {
"type": "string"
},
"slot": {
"description": "slot id, 0 is the first slot",
"type": "integer"
}
}
},
"schema.SlotCacheResponse": {
"type": "object",
"properties": {
"action": {
"description": "\"save\" or \"restore\"",
"type": "string"
},
"duration_ms": {
"description": "time the backend spent",
"type": "number"
},
"id": {
"description": "issued by the save, sent back to restore",
"type": "string"
},
"message": {
"description": "for example the IDs a save pruned",
"type": "string"
},
"model": {
"type": "string"
},
"n_bytes": {
"description": "bytes written or read",
"type": "integer"
},
"n_tokens": {
"description": "tokens saved or restored",
"type": "integer"
},
"slot": {
"type": "integer"
}
}
},
"schema.SoundClassification": {
"type": "object",
"properties": {
"index": {
"type": "integer"
},
"label": {
"type": "string"
},
"score": {
"type": "number"
}
}
},
"schema.SoundClassificationResult": {
"type": "object",
"properties": {
"detections": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.SoundClassification"
}
},
"model": {
"type": "string"
}
}
},
"schema.SpeakerEncoder": {
"type": "object",
"properties": {
"dimension": {
"type": "integer"
},
"family": {
"description": "Family is the embedding space of the encoder. The server fills it from the\nloaded encoder; exported profiles do not carry it and it is not matched.",
"type": "string"
},
"identity": {
"type": "string"
}
}
},
"schema.SpeakerProfile": {
"type": "object",
"properties": {
"clean_duration": {
"type": "number"
},
"embedding": {
"type": "array",
"items": {
"type": "number"
}
},
"intervals": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.SpeakerProfileInterval"
}
},
"speaker": {
"type": "integer"
},
"unavailable_reason": {
"type": "string"
}
}
},
"schema.SpeakerProfileInterval": {
"type": "object",
"properties": {
"end": {
"type": "number"
},
"start": {
"type": "number"
}
}
},
"schema.SpeakerProfiles": {
"type": "object",
"properties": {
"encoder": {
"$ref": "#/definitions/schema.SpeakerEncoder"
},
"speakers": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.SpeakerProfile"
}
},
"version": {
"type": "integer"
}
}
},
"schema.StreamOptions": {
"type": "object",
"properties": {
"include_usage": {
"type": "boolean"
}
}
},
"schema.SysInfoModel": {
"type": "object",
"properties": {
"backend": {
"description": "Backend is the engine serving this model. The loader knows only the ID,\nso it is resolved from the model's config; empty when the model was\nloaded without one (a loose file, or a config since removed).",
"type": "string"
},
"id": {
"type": "string"
},
"process": {
"description": "Process is the backend process serving the model on this host. Absent\nwhen the model has no local process (a distributed worker holds it) or\nthe process could not be read.",
"allOf": [
{
"$ref": "#/definitions/schema.SysInfoProcess"
}
]
},
"size_vram": {
"description": "SizeVRAM is DRM-accounted resident device memory in bytes. Nil means\nthe backend process tree has no complete supported reading.",
"type": "integer"
}
}
},
"schema.SysInfoProcess": {
"type": "object",
"properties": {
"cpu_percent": {
"description": "CPUPercent is the share of the whole host's CPU used since the previous\nreading, 0-100. Absent on the first reading of a process.",
"type": "number"
},
"memory_percent": {
"type": "number"
},
"pid": {
"type": "integer"
},
"rss_bytes": {
"description": "RSSBytes is resident host memory. Weights offloaded to a GPU are not\nin it.",
"type": "integer"
},
"started_at": {
"type": "string"
}
}
},
"schema.SystemInformationResponse": {
"type": "object",
"properties": {
"backends": {
"description": "available backend engines",
"type": "array",
"items": {
"type": "string"
}
},
"loaded_models": {
"description": "currently loaded models",
"type": "array",
"items": {
"$ref": "#/definitions/schema.SysInfoModel"
}
}
}
},
"schema.SystemOneAnswer": {
"type": "object",
"properties": {
"choice": {
"type": "string"
},
"confidence": {
"type": "number"
},
"entities": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.SystemOneEntity"
}
},
"legend": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"noul": {
"type": "number"
},
"probabilities": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
},
"score": {
"type": "number"
},
"type": {
"type": "string"
}
}
},
"schema.SystemOneEntity": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"end": {
"type": "integer"
},
"start": {
"type": "integer"
},
"text": {
"type": "string"
}
}
},
"schema.SystemOnePermuteRequest": {
"type": "object",
"properties": {
"n_perm": {
"type": "integer"
},
"question": {
"type": "string"
},
"request": {
"$ref": "#/definitions/schema.SystemOneRequest"
},
"seed": {
"type": "integer"
}
}
},
"schema.SystemOnePermuteResponse": {
"type": "object",
"properties": {
"argmax_stable": {
"type": "boolean"
},
"runs": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.SystemOnePermuteRun"
}
},
"spread": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
}
}
},
"schema.SystemOnePermuteRun": {
"type": "object",
"properties": {
"choice": {
"type": "string"
},
"latency_ms": {
"type": "number"
},
"order": {
"type": "array",
"items": {
"type": "string"
}
},
"probabilities": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float64"
}
}
}
},
"schema.SystemOneRequest": {
"type": "object"
},
"schema.SystemOneResponse": {
"type": "object",
"properties": {
"answers": {
"type": "object",
"additionalProperties": {
"$ref": "#/definitions/schema.SystemOneAnswer"
}
},
"latency_ms": {
"type": "number"
},
"model": {
"type": "string"
},
"usage": {
"$ref": "#/definitions/schema.SystemOneUsage"
}
}
},
"schema.SystemOneUsage": {
"type": "object",
"properties": {
"input_tokens": {
"type": "integer"
},
"output_tokens": {
"type": "integer"
}
}
},
"schema.TTSRequest": {
"description": "TTS request body",
"type": "object",
"properties": {
"backend": {
"description": "backend engine override",
"type": "string"
},
"input": {
"description": "text input",
"type": "string"
},
"instructions": {
"description": "Instructions is a free-form, per-request style/voice description. It maps to\nthe OpenAI ` + "`" + `instructions` + "`" + ` field and is forwarded to the backend so expressive\nTTS models (e.g. Qwen3-TTS CustomVoice/VoiceDesign) can vary tone or designed\nvoice per request instead of only via the static YAML option.",
"type": "string"
},
"language": {
"description": "(optional) language to use with TTS model",
"type": "string"
},
"model": {
"type": "string"
},
"params": {
"description": "Params carries optional, backend-specific per-request generation parameters\n(LocalAI extension, e.g. Chatterbox exaggeration/cfg_weight/temperature).",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"response_format": {
"description": "(optional) output format",
"type": "string"
},
"sample_rate": {
"description": "(optional) desired output sample rate",
"type": "integer"
},
"speed": {
"description": "Speed is the OpenAI ` + "`" + `speed` + "`" + ` field (0.25-4.0). It is a pointer so an\nexplicit ` + "`" + `\"speed\": 0` + "`" + ` (invalid, rejected with 400) is distinguishable\nfrom an omitted field (left at the backend default). It is normalised\ninto Params[\"speed\"] so it reaches the backend over the same channel as\nthe other per-request generation parameters.",
"type": "number"
},
"stream": {
"description": "(optional) enable streaming TTS",
"type": "boolean"
},
"voice": {
"description": "voice audio file or speaker id",
"type": "string"
}
}
},
"schema.Task": {
"type": "object",
"properties": {
"created_at": {
"type": "string"
},
"cron": {
"description": "Optional cron expression",
"type": "string"
},
"cron_parameters": {
"description": "Parameters to use when executing cron jobs",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"description": {
"description": "Optional description",
"type": "string"
},
"enabled": {
"description": "Can be disabled without deletion",
"type": "boolean"
},
"id": {
"description": "UUID",
"type": "string"
},
"model": {
"description": "Model name (must have MCP config)",
"type": "string"
},
"multimedia_sources": {
"description": "Multimedia sources (for cron jobs)\nURLs to fetch multimedia content from when cron job executes\nEach source can have custom headers for authentication/authorization",
"type": "array",
"items": {
"$ref": "#/definitions/schema.MultimediaSourceConfig"
}
},
"name": {
"description": "User-friendly name",
"type": "string"
},
"prompt": {
"description": "Template prompt (supports Go template .param syntax)",
"type": "string"
},
"updated_at": {
"type": "string"
},
"webhooks": {
"description": "Webhook configuration (for notifications).\nSupports multiple webhook endpoints.\nWebhooks can handle both success and failure cases using template variables:\n.Job (Job object), .Task (Task object), .Result (if successful),\n.Error (if failed), .Status (job status string).",
"type": "array",
"items": {
"$ref": "#/definitions/schema.WebhookConfig"
}
}
}
},
"schema.ThreeDInput": {
"type": "object",
"properties": {
"label": {
"type": "string"
},
"max_bytes": {
"type": "integer"
},
"name": {
"type": "string"
},
"required": {
"type": "boolean"
},
"type": {
"type": "string"
}
}
},
"schema.ThreeDOperation": {
"type": "object",
"properties": {
"endpoint": {
"type": "string"
},
"id": {
"type": "string"
},
"inputs": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ThreeDInput"
}
},
"label": {
"type": "string"
},
"output": {
"type": "string"
},
"parameters": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.ThreeDParameter"
}
}
}
},
"schema.ThreeDParameter": {
"type": "object",
"properties": {
"advanced": {
"type": "boolean"
},
"default": {
"type": "string"
},
"label": {
"type": "string"
},
"max": {
"type": "number"
},
"min": {
"type": "number"
},
"name": {
"type": "string"
},
"options": {
"type": "array",
"items": {
"type": "string"
}
},
"type": {
"type": "string"
}
}
},
"schema.TokenizeRequest": {
"type": "object",
"properties": {
"content": {
"description": "text to tokenize",
"type": "string"
},
"model": {
"type": "string"
}
}
},
"schema.TokenizeResponse": {
"type": "object",
"properties": {
"tokens": {
"description": "token IDs",
"type": "array",
"items": {
"type": "integer"
}
}
}
},
"schema.ToolCall": {
"type": "object",
"properties": {
"function": {
"$ref": "#/definitions/schema.FunctionCall"
},
"id": {
"type": "string"
},
"index": {
"type": "integer"
},
"type": {
"type": "string"
}
}
},
"schema.VADRequest": {
"description": "VAD request body",
"type": "object",
"properties": {
"audio": {
"description": "raw audio samples as float32 PCM",
"type": "array",
"items": {
"type": "number"
}
},
"model": {
"type": "string"
}
}
},
"schema.VideoReference": {
"type": "object",
"properties": {
"audio": {
"description": "optional audio paired with this video",
"type": "string"
},
"fps": {
"description": "frame rate of the source frames; 0 means the backend default",
"type": "integer"
},
"frames": {
"description": "ordered frame images: URL, base64 or data URI",
"type": "array",
"items": {
"type": "string"
}
}
}
},
"schema.VideoRequest": {
"type": "object",
"properties": {
"audio": {
"description": "URL or base64 audio for audio-conditioned generation",
"type": "string"
},
"cfg_scale": {
"description": "classifier-free guidance scale",
"type": "number"
},
"end_image": {
"description": "URL or base64 of the last frame",
"type": "string"
},
"fps": {
"description": "frames per second",
"type": "integer"
},
"height": {
"description": "output height in pixels",
"type": "integer"
},
"input_reference": {
"description": "reference image or video URL",
"type": "string"
},
"model": {
"type": "string"
},
"negative_prompt": {
"description": "things to avoid in the output",
"type": "string"
},
"num_frames": {
"description": "total number of frames to generate",
"type": "integer"
},
"params": {
"description": "backend-specific generation parameters",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"prompt": {
"description": "text description of the video to generate",
"type": "string"
},
"ref_audios": {
"description": "reference audio clips (WAV); exclusive with start_image/end_image",
"type": "array",
"items": {
"type": "string"
}
},
"ref_images": {
"description": "reference images (URL, base64 or data URI); exclusive with start_image/end_image",
"type": "array",
"items": {
"type": "string"
}
},
"ref_videos": {
"description": "reference videos as frame lists; exclusive with start_image/end_image",
"type": "array",
"items": {
"$ref": "#/definitions/schema.VideoReference"
}
},
"response_format": {
"description": "output format (url or b64_json)",
"type": "string"
},
"seconds": {
"description": "duration in seconds (alternative to num_frames)",
"type": "string"
},
"seed": {
"description": "random seed for reproducibility",
"type": "integer"
},
"size": {
"description": "WxH shorthand (e.g. \"512x512\")",
"type": "string"
},
"start_image": {
"description": "URL or base64 of the first frame",
"type": "string"
},
"step": {
"description": "number of diffusion steps",
"type": "integer"
},
"width": {
"description": "output width in pixels",
"type": "integer"
}
}
},
"schema.VoiceAnalysis": {
"type": "object",
"properties": {
"age": {
"type": "number"
},
"dominant_emotion": {
"type": "string"
},
"dominant_gender": {
"type": "string"
},
"emotion": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float32"
}
},
"end": {
"type": "number"
},
"gender": {
"type": "object",
"additionalProperties": {
"type": "number",
"format": "float32"
}
},
"start": {
"type": "number"
}
}
},
"schema.VoiceAnalyzeRequest": {
"type": "object",
"properties": {
"actions": {
"description": "subset of {\"age\",\"gender\",\"emotion\"}",
"type": "array",
"items": {
"type": "string"
}
},
"audio": {
"type": "string"
},
"model": {
"type": "string"
}
}
},
"schema.VoiceAnalyzeResponse": {
"type": "object",
"properties": {
"segments": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.VoiceAnalysis"
}
}
}
},
"schema.VoiceEmbedRequest": {
"type": "object",
"properties": {
"audio": {
"type": "string"
},
"model": {
"type": "string"
}
}
},
"schema.VoiceEmbedResponse": {
"type": "object",
"properties": {
"dim": {
"type": "integer"
},
"embedding": {
"type": "array",
"items": {
"type": "number"
}
},
"model": {
"type": "string"
}
}
},
"schema.VoiceForgetRequest": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"model": {
"type": "string"
},
"store": {
"type": "string"
}
}
},
"schema.VoiceIdentifyMatch": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"distance": {
"type": "number"
},
"id": {
"type": "string"
},
"labels": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"match": {
"type": "boolean"
},
"name": {
"type": "string"
}
}
},
"schema.VoiceIdentifyRequest": {
"type": "object",
"properties": {
"audio": {
"type": "string"
},
"model": {
"type": "string"
},
"store": {
"type": "string"
},
"threshold": {
"type": "number"
},
"top_k": {
"type": "integer"
}
}
},
"schema.VoiceIdentifyResponse": {
"type": "object",
"properties": {
"matches": {
"type": "array",
"items": {
"$ref": "#/definitions/schema.VoiceIdentifyMatch"
}
}
}
},
"schema.VoiceRegisterRequest": {
"type": "object",
"properties": {
"audio": {
"type": "string"
},
"labels": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"model": {
"type": "string"
},
"name": {
"type": "string"
},
"speaker_profiles": {
"$ref": "#/definitions/schema.SpeakerProfiles"
},
"speaker_slot": {
"type": "integer"
},
"store": {
"type": "string"
}
}
},
"schema.VoiceRegisterResponse": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"registered_at": {
"type": "string"
}
}
},
"schema.VoiceVerifyRequest": {
"type": "object",
"properties": {
"anti_spoofing": {
"type": "boolean"
},
"audio1": {
"type": "string"
},
"audio2": {
"type": "string"
},
"model": {
"type": "string"
},
"threshold": {
"type": "number"
}
}
},
"schema.VoiceVerifyResponse": {
"type": "object",
"properties": {
"confidence": {
"type": "number"
},
"distance": {
"type": "number"
},
"model": {
"type": "string"
},
"processing_time_ms": {
"type": "number"
},
"threshold": {
"type": "number"
},
"verified": {
"type": "boolean"
}
}
},
"schema.WebhookConfig": {
"type": "object",
"properties": {
"headers": {
"description": "Custom headers (e.g., Authorization)",
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"method": {
"description": "HTTP method (POST, PUT, PATCH) - default: POST",
"type": "string"
},
"payload_template": {
"description": "Optional template for payload",
"type": "string"
},
"url": {
"description": "Webhook endpoint URL",
"type": "string"
}
}
},
"updates.State": {
"type": "string",
"enum": [
"unknown",
"up-to-date",
"update-available",
"incompatible",
"check-failed",
"integrity-error"
],
"x-enum-varnames": [
"Unknown",
"UpToDate",
"UpdateAvailable",
"Incompatible",
"CheckFailed",
"IntegrityError"
]
},
"updates.Status": {
"type": "object",
"properties": {
"addonId": {
"type": "string",
"example": "notes"
},
"available": {
"description": "Available is true only for update-available.",
"type": "boolean"
},
"candidateDigest": {
"type": "string"
},
"candidateVersion": {
"type": "string",
"example": "1.1.0"
},
"channel": {
"type": "string",
"example": "development"
},
"checkedAt": {
"description": "CheckedAt is when the status was computed. LastSuccess is when the\nlibrary was last read successfully: a failed check keeps it, so the\npage can say how old the knowledge is.",
"type": "string"
},
"error": {
"description": "Error is the redacted class of a failed check.",
"type": "string",
"example": "refresh-failed"
},
"generation": {
"description": "Generation is the installed generation the status describes.",
"type": "integer",
"example": 1
},
"installedDigest": {
"type": "string"
},
"installedVersion": {
"type": "string",
"example": "1.0.0"
},
"lastSuccess": {
"type": "string"
},
"reasons": {
"description": "Reasons are the compatibility reasons of an incompatible or\nintegrity-error status.",
"type": "array",
"items": {
"type": "string"
}
},
"selection": {
"type": "string",
"example": "1.0.0"
},
"sourceId": {
"description": "SourceID, Channel and Selection are the library, channel and\nselection the add-on was installed from. They never change by themselves.",
"type": "string",
"example": "official"
},
"stale": {
"description": "Stale is true when the library data or the cached answer is older than\nthe time to live.",
"type": "boolean"
},
"state": {
"enum": [
"unknown",
"up-to-date",
"update-available",
"incompatible",
"check-failed",
"integrity-error"
],
"allOf": [
{
"$ref": "#/definitions/updates.State"
}
],
"example": "update-available"
},
"throttled": {
"description": "Throttled is true when an explicit refresh was not performed because\none ran moments ago. The status is then the last one.",
"type": "boolean"
}
}
},
"voiceprofile.AudioMetadata": {
"type": "object",
"properties": {
"bit_depth": {
"type": "integer"
},
"channels": {
"type": "integer"
},
"duration_ms": {
"type": "integer"
},
"mime_type": {
"type": "string"
},
"sample_rate": {
"type": "integer"
},
"size_bytes": {
"type": "integer"
}
}
},
"voiceprofile.Profile": {
"type": "object",
"properties": {
"audio": {
"$ref": "#/definitions/voiceprofile.AudioMetadata"
},
"consent_confirmed_at": {
"type": "string"
},
"created_at": {
"type": "string"
},
"description": {
"type": "string"
},
"id": {
"type": "string"
},
"language": {
"type": "string"
},
"name": {
"type": "string"
},
"references": {
"type": "array",
"items": {
"$ref": "#/definitions/voiceprofile.ReferenceMetadata"
}
},
"transcript": {
"type": "string"
},
"updated_at": {
"type": "string"
},
"voice": {
"type": "string"
}
}
},
"voiceprofile.ReferenceMetadata": {
"type": "object",
"properties": {
"audio": {
"$ref": "#/definitions/voiceprofile.AudioMetadata"
},
"transcript": {
"type": "string"
}
}
}
},
"securityDefinitions": {
"BearerAuth": {
"type": "apiKey",
"name": "Authorization",
"in": "header"
}
}
}`
// SwaggerInfo holds exported Swagger Info so clients can modify it
var SwaggerInfo = &swag.Spec{
Version: "2.0.0",
Host: "",
BasePath: "/",
Schemes: []string{"http", "https"},
Title: "LocalAI API",
Description: "The LocalAI Rest API.",
InfoInstanceName: "swagger",
SwaggerTemplate: docTemplate,
LeftDelim: "{{",
RightDelim: "}}",
}
func init() {
swag.Register(SwaggerInfo.InstanceName(), SwaggerInfo)
}