feat(realtime): script-aware clause chunking + streamed-reply fixes

Opt-in pipeline.streaming.clause_chunking splits the streamed LLM reply
into speakable clauses and synthesizes each as soon as it completes,
lowering time-to-first-audio instead of buffering the whole message. The
splitter is script-aware (rivo/uniseg, pure Go): UAX#29 sentence
segmentation handles CJK 。！？ with no whitespace, CJK clause
punctuation (，、；：) and Thai/Lao spaces give finer cuts, and a UAX#14
line-break cap bounds an over-long punctuation-less run. Unlike the old
ASCII .!?/newline segmenter (dropped in 076dcdbe) it does not degrade to
whole-message buffering for CJK/Thai; scripts needing a dictionary
(Khmer/Burmese) stay buffered until a space or end-of-message. Clauses
are synthesized synchronously in the token callback (the LLM keeps
generating into the gRPC stream meanwhile), so audio still starts
mid-generation. Off by default — the whole-message path is unchanged.

Also fix the streamed-reply path and the Talk page:

- Don't swallow streamed autoparser content as reasoning: the
  tokenizer-template path already delivers reasoning-free content via
  ChatDeltas, so prefilling the thinking start token re-tagged it as an
  unclosed reasoning block, leaving no spoken reply. Disable the prefill
  on that path; closed tag pairs are still stripped (#9985).

- Generate collision-free realtime IDs (16 random bytes) instead of a
  constant, so per-item bookkeeping (cancel, conversation.item.retrieve)
  works.

- Key the Talk transcript by the server item_id and upsert entries.
  Realtime events arrive over a WebRTC data channel — outside React's
  event system — so React defers the setTranscript updaters while
  synchronous ref writes in handler bodies run first; the old
  index-tracking ref rendered a duplicate assistant bubble on
  completion. Upserts by item_id are idempotent and order-independent.

- Drop the partial assistant bubble on a cancelled response (barge-in):
  the server discards the interrupted item and sends response.done with
  status "cancelled"; mirror that in the UI so the regenerated reply
  isn't rendered as a second assistant message.

Assisted-by: Claude:claude-opus-4-8 [Claude Code]
Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

.github/backend-matrix.yml

@@ -1766,20 +1766,6 @@ include:
     dockerfile: "./backend/Dockerfile.llama-cpp"
     context: "./"
     ubuntu-version: '2404'
-  - build-type: 'hipblas'
-    cuda-major-version: ""
-    cuda-minor-version: ""
-    platforms: 'linux/amd64'
-    tag-latest: 'auto'
-    tag-suffix: '-gpu-rocm-hipblas-turboquant'
-    builder-base-image: 'quay.io/go-skynet/ci-cache:base-grpc-rocm-amd64'
-    runs-on: 'ubuntu-latest'
-    base-image: "rocm/dev-ubuntu-24.04:7.2.1"
-    skip-drivers: 'false'
-    backend: "turboquant"
-    dockerfile: "./backend/Dockerfile.turboquant"
-    context: "./"
-    ubuntu-version: '2404'
   - build-type: 'hipblas'
     cuda-major-version: ""
     cuda-minor-version: ""

Makefile

@@ -180,7 +180,7 @@ osx-signed: build
 
 ## Run
 run: ## run local-ai
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" $(GOCMD) run ./
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" $(GOCMD) run ./cmd/local-ai
 
 prepare-test: protogen-go build-mock-backend
 

README.md

@@ -149,6 +149,16 @@ local-ai run https://gist.githubusercontent.com/.../phi-2.yaml
 local-ai run oci://localai/phi-2:latest
 ```
 
+To test a running LocalAI server from the terminal, open an interactive chat session from another shell. Inside the prompt, `/models` lists installed models and `/model <name>` switches between them.
+
+```bash
+# Terminal 1
+local-ai run llama-3.2-1b-instruct:q4_k_m
+
+# Terminal 2
+local-ai chat --model llama-3.2-1b-instruct:q4_k_m
+```
+
 > **Automatic Backend Detection**: LocalAI automatically detects your GPU capabilities and downloads the appropriate backend. For advanced options, see [GPU Acceleration](https://localai.io/features/gpu-acceleration/).
 
 For more details, see the [Getting Started guide](https://localai.io/basics/getting_started/).

backend/cpp/ds4/CMakeLists.txt

@@ -60,10 +60,12 @@ elseif(DS4_GPU STREQUAL "cpu")
     set(DS4_OBJS "${DS4_DIR}/ds4_cpu.o")
 endif()
 
-# ds4.c now references ds4_distributed.c (distributed inference was split into
-# its own translation unit upstream). It is a single GPU-agnostic object shared
-# by every GPU mode, so link it in regardless of DS4_GPU.
+# ds4.c now references ds4_distributed.c (distributed inference) and ds4_ssd.c
+# (SSD expert-cache), each split into its own translation unit upstream. Both
+# are GPU-agnostic objects shared by every GPU mode, so link them in regardless
+# of DS4_GPU.
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_distributed.o")
+list(APPEND DS4_OBJS "${DS4_DIR}/ds4_ssd.o")
 
 add_executable(${TARGET}
     grpc-server.cpp

backend/cpp/ds4/Makefile

@@ -1,10 +1,10 @@
 # ds4 backend Makefile.
 #
-# Upstream pin lives below as DS4_VERSION?=477c0e82e2699b35a65fd0a1ed6fe66b41087dfe
+# Upstream pin lives below as DS4_VERSION?=91bafb5acd5a6cf00b1e55ef68bf40ddd207bee7
 # (.github/bump_deps.sh) can find and update it - matches the
 # llama-cpp / ik-llama-cpp / turboquant convention.
 
-DS4_VERSION?=477c0e82e2699b35a65fd0a1ed6fe66b41087dfe
+DS4_VERSION?=91bafb5acd5a6cf00b1e55ef68bf40ddd207bee7
 DS4_REPO?=https://github.com/antirez/ds4
 
 CURRENT_MAKEFILE_DIR := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))
@@ -18,19 +18,20 @@ UNAME_S := $(shell uname -s)
 
 CMAKE_ARGS ?= -DCMAKE_BUILD_TYPE=Release
 
-# ds4_distributed.o is a GPU-agnostic translation unit that ds4.c/ds4_cpu.o now
-# reference (upstream split distributed inference into its own .c). The same
-# object is shared by every GPU mode, so it is appended unconditionally below.
+# ds4_distributed.o and ds4_ssd.o are GPU-agnostic translation units that
+# ds4.c/ds4_cpu.o now reference (upstream split distributed inference and the
+# SSD expert-cache into their own .c files). Both objects are shared by every
+# GPU mode, so they are appended unconditionally below.
 ifeq ($(BUILD_TYPE),cublas)
     CMAKE_ARGS += -DDS4_GPU=cuda
-    DS4_OBJ_TARGET := ds4.o ds4_cuda.o ds4_distributed.o
+    DS4_OBJ_TARGET := ds4.o ds4_cuda.o ds4_distributed.o ds4_ssd.o
 else ifeq ($(UNAME_S),Darwin)
     CMAKE_ARGS += -DDS4_GPU=metal
-    DS4_OBJ_TARGET := ds4.o ds4_metal.o ds4_distributed.o
+    DS4_OBJ_TARGET := ds4.o ds4_metal.o ds4_distributed.o ds4_ssd.o
 else
     # CPU reference path (Linux only - macOS CPU path is broken by VM bug per ds4 README).
     CMAKE_ARGS += -DDS4_GPU=cpu
-    DS4_OBJ_TARGET := ds4_cpu.o ds4_distributed.o
+    DS4_OBJ_TARGET := ds4_cpu.o ds4_distributed.o ds4_ssd.o
 endif
 
 ifneq ($(NATIVE),true)
@@ -55,11 +56,11 @@ ds4:
 # the right per-platform compile flags (Objective-C/Metal on Darwin, nvcc on Linux+CUDA).
 ds4/ds4.o: ds4
 ifeq ($(BUILD_TYPE),cublas)
-	+$(MAKE) -C ds4 ds4.o ds4_cuda.o ds4_distributed.o
+	+$(MAKE) -C ds4 ds4.o ds4_cuda.o ds4_distributed.o ds4_ssd.o
 else ifeq ($(UNAME_S),Darwin)
-	+$(MAKE) -C ds4 ds4.o ds4_metal.o ds4_distributed.o
+	+$(MAKE) -C ds4 ds4.o ds4_metal.o ds4_distributed.o ds4_ssd.o
 else
-	+$(MAKE) -C ds4 ds4_cpu.o ds4_distributed.o
+	+$(MAKE) -C ds4 ds4_cpu.o ds4_distributed.o ds4_ssd.o
 endif
 
 grpc-server: ds4/ds4.o

backend/cpp/ik-llama-cpp/Makefile

@@ -1,5 +1,5 @@
 
-IK_LLAMA_VERSION?=1520eda980564241434b791ce2bbbd128c4be9ea
+IK_LLAMA_VERSION?=e6f8112f3ba126eed3ff5b30cdd08085414a7516
 LLAMA_REPO?=https://github.com/ikawrakow/ik_llama.cpp
 
 CMAKE_ARGS?=

backend/cpp/llama-cpp/Makefile

@@ -1,5 +1,5 @@
 
-LLAMA_VERSION?=7c158fbb4aec1bdc9c81d6ca0e785139f4826fae
+LLAMA_VERSION?=039e20a2db9e87b2477c76cc04905f3e1acad77f
 LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
 
 CMAKE_ARGS?=

backend/cpp/llama-cpp/grpc-server.cpp

@@ -381,6 +381,15 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, const
             });
     }
 
+    // for each video in the request, add the video data
+    for (int i = 0; i < predict->videos_size(); i++) {
+        data["video_data"].push_back(json
+            {
+                {"id", i},
+                {"data",    predict->videos(i)},
+            });
+    }
+
     data["stop"] = predict->stopprompts();
     // data["n_probs"] = predict->nprobs();
     //TODO: images,
@@ -482,23 +491,13 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
     if (!request->draftmodel().empty()) {
         params.speculative.draft.mparams.path = request->draftmodel();
         // Default to draft type if a draft model is set but no explicit type.
-        // Upstream (post ggml-org/llama.cpp#22838) made the speculative type a
-        // vector; the turboquant fork still uses the legacy scalar. The
-        // LOCALAI_LEGACY_LLAMA_CPP_SPEC macro is injected by
-        // backend/cpp/turboquant/patch-grpc-server.sh for fork builds only.
-        // Upstream renamed COMMON_SPECULATIVE_TYPE_DRAFT -> ..._DRAFT_SIMPLE
-        // in ggml-org/llama.cpp#22964; the fork still uses the old name.
-#ifdef LOCALAI_LEGACY_LLAMA_CPP_SPEC
-        if (params.speculative.type == COMMON_SPECULATIVE_TYPE_NONE) {
-            params.speculative.type = COMMON_SPECULATIVE_TYPE_DRAFT;
-        }
-#else
+        // Upstream made the speculative type a vector (ggml-org/llama.cpp#22838)
+        // and renamed COMMON_SPECULATIVE_TYPE_DRAFT -> ..._DRAFT_SIMPLE (#22964).
         const bool no_spec_type = params.speculative.types.empty() ||
             (params.speculative.types.size() == 1 && params.speculative.types[0] == COMMON_SPECULATIVE_TYPE_NONE);
         if (no_spec_type) {
             params.speculative.types = { COMMON_SPECULATIVE_TYPE_DRAFT_SIMPLE };
         }
-#endif
     }
 
     //  params.model_alias ??
@@ -574,9 +573,10 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
     // tokens (0 disables the minimum). Match upstream's default (256). This
     // field was renamed from `checkpoint_every_nt` in llama.cpp; the semantics
     // also shifted from a fixed cadence to a minimum spacing. The turboquant
-    // fork branched before the field existed, so skip it on the legacy path
-    // (LOCALAI_LEGACY_LLAMA_CPP_SPEC is injected by patch-grpc-server.sh).
-#ifndef LOCALAI_LEGACY_LLAMA_CPP_SPEC
+    // fork still lacks common_params::checkpoint_min_step, so skip it there
+    // (LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP is injected by
+    // backend/cpp/turboquant/patch-grpc-server.sh).
+#ifndef LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP
     params.checkpoint_min_step = 256;
 #endif
 
@@ -752,7 +752,7 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
                 params.cache_idle_slots = false;
             }
 
-#ifndef LOCALAI_LEGACY_LLAMA_CPP_SPEC
+#ifndef LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP
         // --- minimum context-checkpoint spacing (upstream -cms / --checkpoint-min-step) ---
         // 0 disables the minimum-spacing gate. Old option names (`checkpoint_every_nt`,
         // `checkpoint_every_n_tokens`) are kept as aliases for backward compatibility
@@ -906,17 +906,6 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
 
         // Speculative decoding options
         } else if (!strcmp(optname, "spec_type") || !strcmp(optname, "speculative_type")) {
-#ifdef LOCALAI_LEGACY_LLAMA_CPP_SPEC
-            // Fork only knows a single scalar `type`. Take the first comma-
-            // separated value and assign it via the singular helper.
-            std::string first = optval_str;
-            const auto comma = first.find(',');
-            if (comma != std::string::npos) first = first.substr(0, comma);
-            auto type = common_speculative_type_from_name(first);
-            if (type != COMMON_SPECULATIVE_TYPE_COUNT) {
-                params.speculative.type = type;
-            }
-#else
             // Upstream switched to a vector of types (comma-separated for multi-type
             // chaining via common_speculative_types_from_names). We keep accepting a
             // single value here, but also tolerate comma-separated lists.
@@ -945,7 +934,6 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
             if (!parsed.empty()) {
                 params.speculative.types = parsed;
             }
-#endif
         } else if (!strcmp(optname, "spec_n_max") || !strcmp(optname, "draft_max")) {
             if (optval != NULL) {
                 try { params.speculative.draft.n_max = std::stoi(optval_str); } catch (...) {}
@@ -983,21 +971,6 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
             // shares the target context size. Accept the option for backward
             // compatibility but silently ignore it.
 
-// Everything below relies on struct shape introduced in ggml-org/llama.cpp#22838
-// (parallel drafting): `ngram_mod`, `ngram_map_k`, `ngram_map_k4v`,
-// `ngram_cache`, and the `draft.{cache_type_*, cpuparams*, tensor_buft_overrides}`
-// fields. The turboquant fork branched before that, so its build defines
-// LOCALAI_LEGACY_LLAMA_CPP_SPEC via patch-grpc-server.sh and these option
-// keys become unrecognized (silently dropped, like any unknown opt) for it.
-//
-// The `#ifdef LOCALAI_LEGACY_LLAMA_CPP_SPEC` / `#else` split below sits at the
-// closing-brace position of the `draft_ctx_size` branch on purpose: in the
-// legacy build the chain ends here (the brace closes draft_ctx_size), and in
-// the modern build the chain continues with `} else if (...)` instead, so the
-// brace count stays balanced under both branches of the preprocessor.
-#ifdef LOCALAI_LEGACY_LLAMA_CPP_SPEC
-        }
-#else
         // --- ngram_mod family (upstream --spec-ngram-mod-*) ---
         } else if (!strcmp(optname, "spec_ngram_mod_n_min")) {
             if (optval != NULL) {
@@ -1127,7 +1100,6 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
             }
             if (!cur.empty()) flush(cur);
         }
-#endif // LOCALAI_LEGACY_LLAMA_CPP_SPEC — closes the `else`/`#ifdef` opened at draft_ctx_size
     }
 
     // Set params.n_parallel from environment variable if not set via options (fallback)
@@ -1177,15 +1149,11 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
             params.tensor_buft_overrides.push_back({nullptr, nullptr});
         }
     }
-    // The draft tensor_buft_overrides are only populated under the modern
-    // (post-#22838) layout, whose population code is itself gated by
-    // LOCALAI_LEGACY_LLAMA_CPP_SPEC above. The turboquant fork lacks
-    // common_params_speculative::draft entirely, so skip the sentinel there too.
-#ifndef LOCALAI_LEGACY_LLAMA_CPP_SPEC
+    // Terminate the draft tensor_buft_overrides list with a sentinel, mirroring
+    // the main-model handling above.
     if (!params.speculative.draft.tensor_buft_overrides.empty()) {
         params.speculative.draft.tensor_buft_overrides.push_back({nullptr, nullptr});
     }
-#endif
 
     // TODO: Add yarn
 
@@ -1544,7 +1512,7 @@ public:
                     msg_json["role"] = msg.role();
 
                     bool is_last_user_msg = (i == last_user_msg_idx);
-                    bool has_images_or_audio = (request->images_size() > 0 || request->audios_size() > 0);
+                    bool has_images_or_audio = (request->images_size() > 0 || request->audios_size() > 0 || request->videos_size() > 0);
 
                     // Handle content - can be string, null, or array
                     // For multimodal content, we'll embed images/audio from separate fields
@@ -1595,6 +1563,16 @@ public:
                                     content_array.push_back(audio_chunk);
                                 }
                             }
+                            if (request->videos_size() > 0) {
+                                for (int j = 0; j < request->videos_size(); j++) {
+                                    json video_chunk;
+                                    video_chunk["type"] = "input_video";
+                                    json input_video;
+                                    input_video["data"] = request->videos(j);
+                                    video_chunk["input_video"] = input_video;
+                                    content_array.push_back(video_chunk);
+                                }
+                            }
                             msg_json["content"] = content_array;
                         } else {
                             // Use content as-is (already array or not last user message)
@@ -1629,6 +1607,16 @@ public:
                                 content_array.push_back(audio_chunk);
                             }
                         }
+                        if (request->videos_size() > 0) {
+                            for (int j = 0; j < request->videos_size(); j++) {
+                                json video_chunk;
+                                video_chunk["type"] = "input_video";
+                                json input_video;
+                                input_video["data"] = request->videos(j);
+                                video_chunk["input_video"] = input_video;
+                                content_array.push_back(video_chunk);
+                            }
+                        }
                         msg_json["content"] = content_array;
                     } else if (msg.role() == "tool") {
                         // Tool role messages must have content field set, even if empty
@@ -2080,6 +2068,16 @@ public:
                         files.push_back(decoded_data);
                     }
                 }
+
+                const auto &video_data = data.find("video_data");
+                if (video_data != data.end() && video_data->is_array())
+                {
+                    for (const auto &video : *video_data)
+                    {
+                        auto decoded_data = base64_decode(video["data"].get<std::string>());
+                        files.push_back(decoded_data);
+                    }
+                }
             }
 
             const bool has_mtmd = ctx_server.impl->mctx != nullptr;
@@ -2332,7 +2330,7 @@ public:
                     }
 
                     bool is_last_user_msg = (i == last_user_msg_idx);
-                    bool has_images_or_audio = (request->images_size() > 0 || request->audios_size() > 0);
+                    bool has_images_or_audio = (request->images_size() > 0 || request->audios_size() > 0 || request->videos_size() > 0);
 
                     // Handle content - can be string, null, or array
                     // For multimodal content, we'll embed images/audio from separate fields
@@ -2385,6 +2383,16 @@ public:
                                     content_array.push_back(audio_chunk);
                                 }
                             }
+                            if (request->videos_size() > 0) {
+                                for (int j = 0; j < request->videos_size(); j++) {
+                                    json video_chunk;
+                                    video_chunk["type"] = "input_video";
+                                    json input_video;
+                                    input_video["data"] = request->videos(j);
+                                    video_chunk["input_video"] = input_video;
+                                    content_array.push_back(video_chunk);
+                                }
+                            }
                             msg_json["content"] = content_array;
                         } else {
                             // Use content as-is (already array or not last user message)
@@ -2424,6 +2432,16 @@ public:
                                 content_array.push_back(audio_chunk);
                             }
                         }
+                        if (request->videos_size() > 0) {
+                            for (int j = 0; j < request->videos_size(); j++) {
+                                json video_chunk;
+                                video_chunk["type"] = "input_video";
+                                json input_video;
+                                input_video["data"] = request->videos(j);
+                                video_chunk["input_video"] = input_video;
+                                content_array.push_back(video_chunk);
+                            }
+                        }
                         msg_json["content"] = content_array;
                         SRV_INF("[CONTENT DEBUG] Predict: Message %d created content array with media\n", i);
                     } else if (!msg.tool_calls().empty()) {
@@ -2886,6 +2904,16 @@ public:
                         files.push_back(decoded_data);
                     }
                 }
+
+                const auto &video_data = data.find("video_data");
+                if (video_data != data.end() && video_data->is_array())
+                {
+                    for (const auto &video : *video_data)
+                    {
+                        auto decoded_data = base64_decode(video["data"].get<std::string>());
+                        files.push_back(decoded_data);
+                    }
+                }
             }
 
             // process files

backend/cpp/turboquant/Makefile

@@ -1,7 +1,7 @@
 
 # Pinned to the HEAD of feature/turboquant-kv-cache on https://github.com/TheTom/llama-cpp-turboquant.
 # Auto-bumped nightly by .github/workflows/bump_deps.yaml.
-TURBOQUANT_VERSION?=5aeb2fdbe26cd4c534c6fa15de73cb5749bd0403
+TURBOQUANT_VERSION?=7d9715f1f071fa07c7b2ad3dbfd320b314139e65
 LLAMA_REPO?=https://github.com/TheTom/llama-cpp-turboquant
 
 CMAKE_ARGS?=

backend/cpp/turboquant/patch-grpc-server.sh

@@ -4,21 +4,19 @@
 #
 #   1. Augment the kv_cache_types[] allow-list so `LoadModel` accepts the
 #      fork-specific `turbo2` / `turbo3` / `turbo4` cache types.
-#   2. Replace `get_media_marker()` (added upstream in ggml-org/llama.cpp#21962,
-#      server-side random per-instance marker) with the legacy "<__media__>"
-#      literal. The fork branched before that PR, so server-common.cpp has no
-#      get_media_marker symbol. The fork's mtmd_default_marker() still returns
-#      "<__media__>", and Go-side tooling falls back to that sentinel when the
-#      backend does not expose media_marker, so substituting the literal keeps
-#      behavior identical on the turboquant path.
-#   3. Revert the `common_params_speculative` field references to the
-#      pre-refactor flat layout. Upstream ggml-org/llama.cpp#22397 split the
-#      struct into nested `draft` / `ngram_simple` / `ngram_mod` / etc. members;
-#      the turboquant fork branched before that PR and still exposes the flat
-#      `n_max`, `mparams_dft`, `ngram_size_n`, ... fields. The substitutions
-#      below map the new nested paths back to the legacy flat names so the
-#      shared grpc-server.cpp keeps compiling against the fork's common.h.
-#      Drop this block once the fork rebases past #22397.
+#   2. Define LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP at the top of the file
+#      so the grpc-server option parser skips the two references to
+#      common_params::checkpoint_min_step (the default and the option handler).
+#      That field does not exist in the fork yet; drop this once it does.
+#
+# The fork used to lag upstream on the whole common_params_speculative refactor
+# (ggml-org/llama.cpp#22397/#22838/#22964), the model_tgt rename (#22838) and
+# get_media_marker (#21962), which required a much larger compat shim here
+# (flat-field sed renames + a coarse LOCALAI_LEGACY_LLAMA_CPP_SPEC define). The
+# fork has since rebased past all of those, so the only remaining gap is
+# checkpoint_min_step. If a future bump reintroduces a divergence, add a narrow
+# guard in grpc-server.cpp keyed on a fork-specific macro and inject it here
+# rather than resurrecting the coarse one.
 #
 # We patch the *copy* sitting in turboquant-<flavor>-build/, never the original
 # under backend/cpp/llama-cpp/, so the stock llama-cpp build keeps compiling
@@ -72,72 +70,20 @@ else
     echo "==> KV allow-list patch OK"
 fi
 
-if grep -q 'get_media_marker()' "$SRC"; then
-    echo "==> patching $SRC to replace get_media_marker() with legacy \"<__media__>\" literal"
-    # Only one call site today (ModelMetadata), but replace all occurrences to
-    # stay robust if upstream adds more. Use a temp file to avoid relying on
-    # sed -i portability (the builder image uses GNU sed, but keeping this
-    # consistent with the awk block above).
-    sed 's/get_media_marker()/"<__media__>"/g' "$SRC" > "$SRC.tmp"
-    mv "$SRC.tmp" "$SRC"
-    echo "==> get_media_marker() substitution OK"
+# 2. Define LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP at the top of the file so
+#    the grpc-server option parser skips the two references to
+#    common_params::checkpoint_min_step (the default assignment and the option
+#    handler). That field does not exist in the fork yet. Drop this block once
+#    the fork rebases past the bump that added checkpoint_min_step.
+if grep -q '^#define LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP' "$SRC"; then
+    echo "==> $SRC already defines LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP, skipping"
 else
-    echo "==> $SRC has no get_media_marker() call, skipping media-marker patch"
-fi
-
-if grep -q 'params\.speculative\.draft\.\|params\.speculative\.ngram_simple\.' "$SRC"; then
-    echo "==> patching $SRC to revert common_params_speculative refs to pre-#22397 flat layout"
-    # Each substitution is the exact post-refactor path → legacy flat field.
-    # Order doesn't matter because the source paths are disjoint, but we keep
-    # the most-specific (mparams.path) first for readability.
-    sed -E \
-        -e 's/params\.speculative\.draft\.mparams\.path/params.speculative.mparams_dft.path/g' \
-        -e 's/params\.speculative\.draft\.n_max/params.speculative.n_max/g' \
-        -e 's/params\.speculative\.draft\.n_min/params.speculative.n_min/g' \
-        -e 's/params\.speculative\.draft\.p_min/params.speculative.p_min/g' \
-        -e 's/params\.speculative\.draft\.p_split/params.speculative.p_split/g' \
-        -e 's/params\.speculative\.draft\.n_gpu_layers/params.speculative.n_gpu_layers/g' \
-        -e 's/params\.speculative\.draft\.n_ctx/params.speculative.n_ctx/g' \
-        -e 's/params\.speculative\.ngram_simple\.size_n/params.speculative.ngram_size_n/g' \
-        -e 's/params\.speculative\.ngram_simple\.size_m/params.speculative.ngram_size_m/g' \
-        -e 's/params\.speculative\.ngram_simple\.min_hits/params.speculative.ngram_min_hits/g' \
-        "$SRC" > "$SRC.tmp"
-    mv "$SRC.tmp" "$SRC"
-    echo "==> speculative field rename OK"
-else
-    echo "==> $SRC has no post-#22397 speculative field refs, skipping spec rename patch"
-fi
-
-# 4. Revert the `ctx_server.impl->model_tgt` rename introduced by upstream
-#    ggml-org/llama.cpp#22838 (parallel drafting). The turboquant fork still
-#    exposes the field as `model` on `server_context_impl`. The two call sites
-#    are in the Rerank and ModelMetadata RPC handlers.
-if grep -q 'ctx_server\.impl->model_tgt' "$SRC"; then
-    echo "==> patching $SRC to revert ctx_server.impl->model_tgt -> ctx_server.impl->model"
-    sed -E 's/ctx_server\.impl->model_tgt/ctx_server.impl->model/g' "$SRC" > "$SRC.tmp"
-    mv "$SRC.tmp" "$SRC"
-    echo "==> model_tgt rename OK"
-else
-    echo "==> $SRC has no ctx_server.impl->model_tgt refs, skipping model_tgt rename patch"
-fi
-
-# 5. Define LOCALAI_LEGACY_LLAMA_CPP_SPEC at the top of the file so the
-#    grpc-server option parser skips the new option-handler blocks (ngram_mod,
-#    ngram_map_k, ngram_map_k4v, ngram_cache, draft.cache_type_*, draft.cpuparams*,
-#    draft.tensor_buft_overrides) introduced for the post-#22838 layout, the
-#    draft.tensor_buft_overrides sentinel termination, and the
-#    common_params::checkpoint_min_step default/option (added with the
-#    35c9b1f3 bump). Those blocks reference struct fields that simply do not
-#    exist in the fork.
-if grep -q '^#define LOCALAI_LEGACY_LLAMA_CPP_SPEC' "$SRC"; then
-    echo "==> $SRC already defines LOCALAI_LEGACY_LLAMA_CPP_SPEC, skipping"
-else
-    echo "==> patching $SRC to define LOCALAI_LEGACY_LLAMA_CPP_SPEC at the top"
-    # Insert the define before the very first `#include` so it precedes all the
-    # speculative-decoding code paths.
+    echo "==> patching $SRC to define LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP at the top"
+    # Insert the define before the very first `#include` so it precedes the
+    # checkpoint_min_step references.
     awk '
         !done && /^#include/ {
-            print "#define LOCALAI_LEGACY_LLAMA_CPP_SPEC 1"
+            print "#define LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP 1"
             print "// ^ injected by backend/cpp/turboquant/patch-grpc-server.sh"
             print ""
             done = 1
@@ -145,13 +91,13 @@ else
         { print }
         END {
             if (!done) {
-                print "patch-grpc-server.sh: no #include anchor found to insert LOCALAI_LEGACY_LLAMA_CPP_SPEC" > "/dev/stderr"
+                print "patch-grpc-server.sh: no #include anchor found to insert LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP" > "/dev/stderr"
                 exit 1
             }
         }
     ' "$SRC" > "$SRC.tmp"
     mv "$SRC.tmp" "$SRC"
-    echo "==> LOCALAI_LEGACY_LLAMA_CPP_SPEC define OK"
+    echo "==> LOCALAI_TURBOQUANT_NO_CHECKPOINT_MIN_STEP define OK"
 fi
 
 echo "==> all patches applied"

backend/cpp/turboquant/patches/0001-hip-guard-copy2d-peer-fastpath.patch

@@ -0,0 +1,55 @@
+hip: port the turboquant CUDA additions that ggml's HIP shim doesn't cover
+
+The turboquant fork adds/modifies a few ggml-cuda.cu spots with CUDA APIs
+that ggml's HIP (and MUSA) compatibility layer does not provide, breaking
+the -gpu-rocm-hipblas-turboquant build:
+
+  1. ggml_cuda_copy2d_across_devices() (host-staged cross-device copy for
+     split mul_mat output) uses the CUDA 3D-peer copy APIs
+     cudaMemcpy3DPeerParms / make_cudaPitchedPtr / make_cudaExtent /
+     cudaMemcpy3DPeerAsync. HIP genuinely does not support these (see the
+     fork's own comment "HIP does not support cudaMemcpy3DPeerAsync"), so
+     guard the peer fast path with #if !defined(GGML_USE_HIP) &&
+     !defined(GGML_USE_MUSA) -- matching how the fork already guards the
+     same API for the sibling 2D copy -- and fall through to the existing
+     cudaMemcpyAsync staging fallback below (functionally identical,
+     slightly slower on multi-GPU ROCm).
+
+  2. ggml_backend_cuda_device_event_new() creates its event with plain
+     cudaEventCreate, which ggml's HIP shim does not alias (it only aliases
+     cudaEventCreateWithFlags). Use cudaEventCreateWithFlags(..., 
+     cudaEventDisableTiming) -- exactly what the rest of this file already
+     does (cf. lines ~1034, ~3461) and HIP-safe.
+
+CUDA builds are unaffected. Drop the relevant hunk once the fork HIP-ports
+these; apply-patches.sh fails fast if an anchor goes stale.
+
+diff --git a/ggml/src/ggml-cuda/ggml-cuda.cu b/ggml/src/ggml-cuda/ggml-cuda.cu
+index 0427e6b..6352e6a 100644
+--- a/ggml/src/ggml-cuda/ggml-cuda.cu
++++ b/ggml/src/ggml-cuda/ggml-cuda.cu
+@@ -1933,6 +1933,7 @@ static cudaError_t ggml_cuda_copy2d_across_devices(
+     size_t width, size_t height, cudaStream_t dst_stream, cudaStream_t src_stream) {
+ 
+     const auto & info = ggml_cuda_info();
++#if !defined(GGML_USE_HIP) && !defined(GGML_USE_MUSA)  // 3D-peer copy types unmapped by ggml's HIP/MUSA shim; use staging fallback below
+     if (info.peer_access[src_device][dst_device]) {
+         cudaMemcpy3DPeerParms p = {};
+         p.dstDevice = dst_device;
+@@ -1942,6 +1943,7 @@ static cudaError_t ggml_cuda_copy2d_across_devices(
+         p.extent = make_cudaExtent(width, height, 1);
+         return cudaMemcpy3DPeerAsync(&p, dst_stream);
+     }
++#endif // !defined(GGML_USE_HIP) && !defined(GGML_USE_MUSA)
+ 
+     // Fallback: stage all rows through a single contiguous pinned buffer
+     int prev_device = ggml_cuda_get_device();
+@@ -5714,7 +5716,7 @@ static ggml_backend_event_t ggml_backend_cuda_device_event_new(ggml_backend_dev_
+     ggml_cuda_set_device(dev_ctx->device);
+ 
+     cudaEvent_t event;
+-    CUDA_CHECK(cudaEventCreate(&event));
++    CUDA_CHECK(cudaEventCreateWithFlags(&event, cudaEventDisableTiming));
+ 
+     return new ggml_backend_event {
+         /* .device  = */ dev,

backend/go/crispasr/CMakeLists.txt

@@ -14,7 +14,7 @@ target_include_directories(gocrispasr PRIVATE
 # whisper. crispasr is the referencer; the backend static libs supply the
 # per-architecture symbols; ggml is the math/runtime base.
 target_link_libraries(gocrispasr PRIVATE
-    crispasr
+    crispasr-lib
     parakeet canary canary_ctc cohere granite_speech granite_nle
     voxtral voxtral4b qwen3_asr qwen3_tts orpheus chatterbox indextts
     kokoro voxcpm2_tts m2m100 t5_translate wav2vec2-ggml vibevoice

backend/go/crispasr/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # CrispASR version (release tag)
 CRISPASR_REPO?=https://github.com/CrispStrobe/CrispASR
-CRISPASR_VERSION?=13d54e110e1538e0f0bc3af0680b9ab246cfb48d
+CRISPASR_VERSION?=c29f6653a516a3001d923944dad8892072cc7334
 SO_TARGET?=libgocrispasr.so
 
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF

backend/go/parakeet-cpp/Makefile

@@ -1,6 +1,6 @@
 # parakeet-cpp backend Makefile.
 #
-# Upstream pin lives below as PARAKEET_VERSION?=b11fe5bca78ad8b342dd559a43d76df3984bb447
+# Upstream pin lives below as PARAKEET_VERSION?=e270af73b94c9a5c37ec516230219ed4580e1db6
 # (.github/bump_deps.sh) can find and update it - matches the
 # whisper.cpp / ds4 / vibevoice-cpp convention.
 #
@@ -15,7 +15,7 @@
 # That's what the L0 smoke test uses. The default target below does the
 # proper clone-at-pin + cmake build so CI doesn't need a side-checkout.
 
-PARAKEET_VERSION?=b11fe5bca78ad8b342dd559a43d76df3984bb447
+PARAKEET_VERSION?=e270af73b94c9a5c37ec516230219ed4580e1db6
 PARAKEET_REPO?=https://github.com/mudler/parakeet.cpp
 
 GOCMD?=go

backend/go/parakeet-cpp/batcher.go

@@ -7,8 +7,12 @@ import "time"
 type batchRequest struct {
 	pcm     []float32
 	decoder int32
-	tag     string
-	reply   chan batchReply
+	// language is the per-request target locale ("" means the model default).
+	// parakeet.cpp's batched C-API takes ONE target_lang for the whole batch,
+	// so the dispatcher only coalesces requests that share a language.
+	language string
+	tag      string
+	reply    chan batchReply
 }
 
 // batchReply carries one per-item JSON object string (an element of the C-API's
@@ -43,13 +47,25 @@ func newBatcher(maxSize int, maxWait time.Duration, runBatch func([]*batchReques
 // run is the dispatcher loop: accumulate submitted requests until either maxSize
 // is reached or maxWait elapses since the first queued request, then dispatch.
 // Exits when stop is closed (draining any partially-filled batch first).
+//
+// A batch carries ONE language (parakeet.cpp's batched C-API takes a single
+// target_lang), so a request whose language differs from the batch leader is
+// not coalesced: it is held in carry and becomes the leader of the next batch.
+// carry is therefore never dropped and its caller never deadlocks: every batch
+// (including a lone carry on stop) is dispatched, and runBatch replies to all.
 func (b *batcher) run(stop <-chan struct{}) {
+	var carry *batchRequest
 	for {
 		var first *batchRequest
-		select {
-		case first = <-b.submit:
-		case <-stop:
-			return
+		if carry != nil {
+			// A mismatched request from the previous fill leads this batch.
+			first, carry = carry, nil
+		} else {
+			select {
+			case first = <-b.submit:
+			case <-stop:
+				return
+			}
 		}
 		batch := []*batchRequest{first}
 
@@ -64,12 +80,22 @@ func (b *batcher) run(stop <-chan struct{}) {
 		for len(batch) < b.maxSize {
 			select {
 			case r := <-b.submit:
+				if r.language != first.language {
+					// Different language: carry it to the next batch so this
+					// batch stays single-language, then dispatch what we have.
+					carry = r
+					break fill
+				}
 				batch = append(batch, r)
 			case <-timer.C:
 				break fill
 			case <-stop:
 				timer.Stop()
 				b.runBatch(batch)
+				// Don't strand a carried request's caller on shutdown.
+				if carry != nil {
+					b.runBatch([]*batchRequest{carry})
+				}
 				return
 			}
 		}

backend/go/parakeet-cpp/batcher_test.go

@@ -105,4 +105,60 @@ var _ = Describe("batcher", func() {
 		go func() { <-rep }()
 		Eventually(dispatched, "2s").Should(Receive(Equal(1)))
 	})
+
+	It("never coalesces requests with different languages into one batch", func() {
+		// parakeet.cpp's batched C-API takes ONE target_lang per batch, so the
+		// dispatcher must keep every dispatched batch single-language. Submit a
+		// mix of languages and assert (a) no batch ever carries more than one
+		// distinct language and (b) every submitted request still gets a reply
+		// (the mismatched carry-over is never dropped).
+		var mu sync.Mutex
+		var langsPerBatch [][]string
+		run := func(reqs []*batchRequest) {
+			seen := map[string]struct{}{}
+			var distinct []string
+			for _, r := range reqs {
+				if _, ok := seen[r.language]; !ok {
+					seen[r.language] = struct{}{}
+					distinct = append(distinct, r.language)
+				}
+			}
+			mu.Lock()
+			langsPerBatch = append(langsPerBatch, distinct)
+			mu.Unlock()
+			echoReply(reqs)
+		}
+		// Large window + size so the fill loop stays open across submits and the
+		// language constraint (not the timer) is what splits the batches.
+		b := newBatcher(16, 200*time.Millisecond, run)
+		stop := make(chan struct{})
+		go b.run(stop)
+		defer close(stop)
+
+		langs := []string{"en", "en", "de", "de", "en", "fr", "fr"}
+		const N = 7
+		var wg sync.WaitGroup
+		got := make([]string, N)
+		for i := 0; i < N; i++ {
+			wg.Add(1)
+			go func(i int) {
+				defer wg.Done()
+				rep := make(chan batchReply, 1)
+				b.submit <- &batchRequest{tag: string(rune('a' + i)), language: langs[i], reply: rep}
+				got[i] = (<-rep).json
+			}(i)
+		}
+		wg.Wait()
+
+		mu.Lock()
+		defer mu.Unlock()
+		// Invariant: every dispatched batch is single-language.
+		for _, distinct := range langsPerBatch {
+			Expect(len(distinct)).To(Equal(1), "a batch coalesced more than one language: %v", distinct)
+		}
+		// Liveness: every request got a reply (carry-over never stranded).
+		for i := 0; i < N; i++ {
+			Expect(got[i]).To(Equal(string(rune('a' + i))))
+		}
+	})
 })

backend/go/parakeet-cpp/goparakeetcpp.go

@@ -48,6 +48,13 @@ var (
 	// side reads them as const float*/const int*.
 	CppTranscribePcmBatchJSON func(ctx uintptr, samplesConcat []float32, nSamples []int32, nClips int32, sampleRate int32, decoder int32) uintptr
 
+	// CppTranscribePcmBatchJSONLang is the multilingual variant of the batched
+	// JSON entry point: identical, plus a trailing target_lang. "" (the model
+	// default, "auto") is passed for non-prompt models, which ignore it; an
+	// unknown locale on a prompt model returns 0 and sets last_error. Present
+	// only in newer libparakeet.so; nil falls back to CppTranscribePcmBatchJSON.
+	CppTranscribePcmBatchJSONLang func(ctx uintptr, samplesConcat []float32, nSamples []int32, nClips int32, sampleRate int32, decoder int32, targetLang string) uintptr
+
 	// Cache-aware streaming (RNN-T) entry points. stream_begin returns 0 for
 	// non-streaming models. feed/finalize return a malloc'd char* (uintptr,
 	// freed via CppFreeString); feed writes 1 to *eouOut on an <EOU>/<EOB>.
@@ -55,6 +62,18 @@ var (
 	CppStreamFeed     func(s uintptr, pcm []float32, nSamples int32, eouOut unsafe.Pointer) uintptr
 	CppStreamFinalize func(s uintptr) uintptr
 	CppStreamFree     func(s uintptr)
+
+	// CppStreamBeginLang is the multilingual variant of stream_begin: identical,
+	// plus a trailing target_lang ("" means the model default). Present only in
+	// newer libparakeet.so; nil falls back to CppStreamBegin.
+	CppStreamBeginLang func(ctx uintptr, targetLang string) uintptr
+
+	// Streaming JSON variants (ABI v4): feed/finalize returning a malloc'd char*
+	// JSON document {text,eou,frame_sec,words} (uintptr, freed via CppFreeString)
+	// so streaming segments can carry per-word timestamps. Present only in newer
+	// libparakeet.so; nil falls back to the text-only CppStreamFeed/Finalize path.
+	CppStreamFeedJSON     func(s uintptr, pcm []float32, nSamples int32) uintptr
+	CppStreamFinalizeJSON func(s uintptr) uintptr
 )
 
 // streamChunkSamples is how much 16 kHz mono PCM we hand to stream_feed per
@@ -72,9 +91,26 @@ const streamChunkSamples = 16000
 //
 // "start"/"end"/"t" are seconds; "conf" is confidence in (0,1].
 type transcriptJSON struct {
-	Text   string            `json:"text"`
-	Words  []transcriptWord  `json:"words"`
-	Tokens []transcriptToken `json:"tokens"`
+	Text     string            `json:"text"`
+	FrameSec float64           `json:"frame_sec"`
+	Words    []transcriptWord  `json:"words"`
+	Tokens   []transcriptToken `json:"tokens"`
+}
+
+// streamFeedJSON mirrors the document returned by
+// parakeet_capi_stream_feed_json / parakeet_capi_stream_finalize_json (ABI v4):
+//
+//	{"text":"...","eou":0,"frame_sec":0.080000,
+//	 "words":[{"w":"...","start":0.480,"end":0.640,"conf":0.9100}, ...]}
+//
+// "text" is the newly-finalized text since the last call; "eou" is 1 when an
+// <EOU>/<EOB> fired this feed; "words" are the words finalized this call with
+// absolute (stream-relative) start/end seconds.
+type streamFeedJSON struct {
+	Text     string           `json:"text"`
+	Eou      int              `json:"eou"`
+	FrameSec float64          `json:"frame_sec"`
+	Words    []transcriptWord `json:"words"`
 }
 
 type transcriptWord struct {
@@ -103,6 +139,10 @@ type ParakeetCpp struct {
 	engineMu sync.Mutex // sole guard of the one C engine (dispatcher + streaming)
 	bat      *batcher
 	batStop  chan struct{}
+	// segmentGapFrames is NeMo's segment_gap_threshold in ENCODER FRAMES (model
+	// YAML option, default 0=off). When >0 it adds NeMo's silence-gap split on
+	// top of the punctuation split; converted to seconds via the JSON frame_sec.
+	segmentGapFrames int
 }
 
 // Load is the LocalAI gRPC entry point for LoadModel: it calls
@@ -132,6 +172,11 @@ func (p *ParakeetCpp) Load(opts *pb.ModelOptions) error {
 	if maxWaitMs < 0 {
 		maxWaitMs = 0
 	}
+
+	// NeMo's segment_gap_threshold (encoder frames, default 0=off). Off by
+	// default matches NeMo's default (punctuation-only segments); when set it
+	// additionally splits segments on inter-word silence (see transcriptResultFromDoc).
+	p.segmentGapFrames = optInt(opts, "segment_gap_threshold", 0)
 	if CppTranscribePcmBatchJSON != nil {
 		p.batStop = make(chan struct{})
 		p.bat = newBatcher(maxSize, time.Duration(maxWaitMs)*time.Millisecond, p.runBatch)
@@ -187,8 +232,19 @@ func (p *ParakeetCpp) runBatch(reqs []*batchRequest) {
 	if len(reqs) > 0 {
 		dec = reqs[0].decoder
 	}
+	// All requests in a batch share one language (the batcher coalesces only
+	// same-language requests), so any element's language describes the batch.
+	lang := ""
+	if len(reqs) > 0 {
+		lang = reqs[0].language
+	}
 	p.engineMu.Lock()
-	cstr := CppTranscribePcmBatchJSON(p.ctxPtr, concat, nSamples, int32(len(reqs)), 16000, dec)
+	var cstr uintptr
+	if CppTranscribePcmBatchJSONLang != nil {
+		cstr = CppTranscribePcmBatchJSONLang(p.ctxPtr, concat, nSamples, int32(len(reqs)), 16000, dec, lang)
+	} else {
+		cstr = CppTranscribePcmBatchJSON(p.ctxPtr, concat, nSamples, int32(len(reqs)), 16000, dec)
+	}
 	p.engineMu.Unlock()
 	if cstr == 0 {
 		err := fmt.Errorf("parakeet-cpp: batch transcribe failed: %s", CppLastError(p.ctxPtr))
@@ -226,8 +282,9 @@ func (p *ParakeetCpp) runBatch(reqs []*batchRequest) {
 // OpenAI API, whose default is segment-level); token ids always populate
 // Segment.Tokens.
 //
-// translate/diarize/prompt/temperature/language/threads are not applicable to
-// parakeet and are ignored; streaming is handled by AudioTranscriptionStream
+// translate/diarize/prompt/temperature/threads are not applicable to parakeet
+// and are ignored; language is honored on the batched + streaming paths (see
+// opts.GetLanguage() below); streaming is handled by AudioTranscriptionStream
 // (L2).
 func (p *ParakeetCpp) AudioTranscription(ctx context.Context, opts *pb.TranscriptRequest) (pb.TranscriptResult, error) {
 	if p.ctxPtr == 0 {
@@ -259,7 +316,7 @@ func (p *ParakeetCpp) AudioTranscription(ctx context.Context, opts *pb.Transcrip
 		if err := json.Unmarshal([]byte(raw), &doc); err != nil {
 			return pb.TranscriptResult{}, fmt.Errorf("parakeet-cpp: decode transcript json: %w", err)
 		}
-		return transcriptResultFromDoc(doc, opts), nil
+		return transcriptResultFromDoc(doc, opts, p.segmentGapFrames), nil
 	}
 
 	// Batched path: decode to PCM, submit to the batcher, wait for this request's
@@ -271,7 +328,7 @@ func (p *ParakeetCpp) AudioTranscription(ctx context.Context, opts *pb.Transcrip
 	}
 	rep := make(chan batchReply, 1)
 	select {
-	case p.bat.submit <- &batchRequest{pcm: pcm, decoder: 0, reply: rep}:
+	case p.bat.submit <- &batchRequest{pcm: pcm, decoder: 0, language: opts.GetLanguage(), reply: rep}:
 	case <-ctx.Done():
 		return pb.TranscriptResult{}, status.Error(codes.Canceled, "transcription cancelled")
 	}
@@ -288,34 +345,169 @@ func (p *ParakeetCpp) AudioTranscription(ctx context.Context, opts *pb.Transcrip
 	if err := json.Unmarshal([]byte(res.json), &doc); err != nil {
 		return pb.TranscriptResult{}, fmt.Errorf("parakeet-cpp: decode transcript json: %w", err)
 	}
-	return transcriptResultFromDoc(doc, opts), nil
+	return transcriptResultFromDoc(doc, opts, p.segmentGapFrames), nil
 }
 
+// segmentSeparators is NeMo's default segment_seperators (sentence-ending
+// punctuation). Splitting on these matches NeMo's default segment timestamps.
+var segmentSeparators = []rune{'.', '?', '!'}
+
 // transcriptResultFromDoc maps a decoded transcriptJSON to a TranscriptResult,
-// synthesising a single whole-clip segment and attaching word timings only when
-// the caller requested word granularity. Shared by the batched and direct paths.
-func transcriptResultFromDoc(doc transcriptJSON, opts *pb.TranscriptRequest) pb.TranscriptResult {
+// grouping words into NeMo-faithful segments (see splitWordsIntoSegments). The
+// optional gapFrames (NeMo's segment_gap_threshold, in encoder FRAMES; 0=off)
+// additionally splits on inter-word silence; it is converted to a seconds gap
+// with the document's frame_sec. Per-segment word timings are attached only when
+// the caller requested word granularity; token ids populate each segment's
+// Tokens by time-window membership. Shared by the batched and direct paths.
+func transcriptResultFromDoc(doc transcriptJSON, opts *pb.TranscriptRequest, gapFrames int) pb.TranscriptResult {
 	text := strings.TrimSpace(doc.Text)
-	words := make([]*pb.TranscriptWord, 0, len(doc.Words))
-	for _, w := range doc.Words {
-		words = append(words, &pb.TranscriptWord{Start: secondsToNanos(w.Start), End: secondsToNanos(w.End), Text: w.W})
+
+	// Frame-unit gap threshold -> seconds (NeMo segment_gap_threshold). 0 = off.
+	gapSeconds := 0.0
+	if gapFrames > 0 {
+		if doc.FrameSec > 0 {
+			gapSeconds = float64(gapFrames) * doc.FrameSec
+		} else {
+			xlog.Warn("parakeet-cpp: segment_gap_threshold set but libparakeet.so " +
+				"did not report frame_sec; falling back to punctuation-only segments")
+		}
 	}
-	tokens := make([]int32, 0, len(doc.Tokens))
-	for _, t := range doc.Tokens {
-		tokens = append(tokens, t.ID)
+
+	groups := splitWordsIntoSegments(doc.Words, segmentSeparators, gapSeconds)
+	if len(groups) == 0 {
+		// No words (edge case): single whole-clip text segment.
+		return pb.TranscriptResult{
+			Text:     text,
+			Segments: []*pb.TranscriptSegment{{Id: 0, Text: text}},
+		}
 	}
-	var segStart, segEnd int64
-	if len(words) > 0 {
-		segStart = words[0].Start
-		segEnd = words[len(words)-1].End
+
+	wantWords := wordsRequested(opts.TimestampGranularities)
+	segments := make([]*pb.TranscriptSegment, 0, len(groups))
+	for id, group := range groups {
+		parts := make([]string, len(group))
+		for i, gw := range group {
+			parts[i] = gw.W
+		}
+		seg := &pb.TranscriptSegment{
+			Id:     int32(id),
+			Start:  secondsToNanos(group[0].Start),
+			End:    secondsToNanos(group[len(group)-1].End),
+			Text:   strings.TrimSpace(strings.Join(parts, " ")),
+			Tokens: tokensInWindow(doc.Tokens, group[0].Start, group[len(group)-1].End),
+		}
+		if wantWords {
+			ws := make([]*pb.TranscriptWord, len(group))
+			for i, gw := range group {
+				ws[i] = &pb.TranscriptWord{Start: secondsToNanos(gw.Start), End: secondsToNanos(gw.End), Text: gw.W}
+			}
+			seg.Words = ws
+		}
+		segments = append(segments, seg)
 	}
-	seg := &pb.TranscriptSegment{Id: 0, Start: segStart, End: segEnd, Text: text, Tokens: tokens}
-	if wordsRequested(opts.TimestampGranularities) {
-		seg.Words = words
-	}
-	return pb.TranscriptResult{Text: text, Segments: []*pb.TranscriptSegment{seg}}
+	return pb.TranscriptResult{Text: text, Segments: segments}
 }
 
+// splitWordsIntoSegments groups words into segments exactly as NeMo's
+// get_segment_offsets does (nemo/collections/asr/parts/utils/timestamp_utils.py).
+// Walking the words, it closes a segment when (1) the gap rule is enabled
+// (gapSeconds > 0) and the segment already has words and the gap from the
+// previous word's end to this word's start is >= gapSeconds - the current word
+// then STARTS a new segment - or, checked only when the gap rule did not apply
+// (NeMo's elif), (2) the word ends with (or is) a separator, which closes the
+// segment INCLUDING that word. Trailing words flush into a final segment.
+// gapSeconds <= 0 disables the gap rule, matching NeMo's default
+// segment_gap_threshold=None (punctuation-only segments).
+func splitWordsIntoSegments(words []transcriptWord, separators []rune, gapSeconds float64) [][]transcriptWord {
+	var segments [][]transcriptWord
+	var cur []transcriptWord
+	for i, word := range words {
+		gapActive := gapSeconds > 0 && len(cur) > 0
+		if gapActive && (word.Start-words[i-1].End) >= gapSeconds {
+			segments = append(segments, cur)
+			cur = []transcriptWord{word}
+			continue
+		}
+		if !gapActive && endsWithSeparator(word.W, separators) {
+			cur = append(cur, word)
+			segments = append(segments, cur)
+			cur = nil
+			continue
+		}
+		cur = append(cur, word)
+	}
+	if len(cur) > 0 {
+		segments = append(segments, cur)
+	}
+	return segments
+}
+
+// endsWithSeparator reports whether w's last rune is in separators (matching
+// NeMo's `word[-1] in delims or word in delims`).
+func endsWithSeparator(w string, separators []rune) bool {
+	r := []rune(strings.TrimSpace(w))
+	if len(r) == 0 {
+		return false
+	}
+	last := r[len(r)-1]
+	for _, s := range separators {
+		if last == s {
+			return true
+		}
+	}
+	return false
+}
+
+// tokensInWindow returns the ids of tokens whose timestamp t falls in
+// [start, end] (inclusive), assigning each token to the segment that spans its
+// time. The last segment's end is the last word end, so the final token is
+// included.
+func tokensInWindow(tokens []transcriptToken, start, end float64) []int32 {
+	var ids []int32
+	for _, t := range tokens {
+		if t.T >= start && t.T <= end {
+			ids = append(ids, t.ID)
+		}
+	}
+	return ids
+}
+
+// streamSegmenter accumulates streaming words into per-utterance segments. EOU
+// is the model's own utterance boundary; each closed segment takes its start/end
+// from its first/last accumulated word.
+type streamSegmenter struct {
+	segs   []*pb.TranscriptSegment
+	cur    []transcriptWord
+	nextID int32
+}
+
+func (s *streamSegmenter) add(doc streamFeedJSON) {
+	s.cur = append(s.cur, doc.Words...)
+	if doc.Eou != 0 {
+		s.flush()
+	}
+}
+
+func (s *streamSegmenter) flush() {
+	if len(s.cur) == 0 {
+		return
+	}
+	parts := make([]string, len(s.cur))
+	for i, w := range s.cur {
+		parts[i] = w.W
+	}
+	s.segs = append(s.segs, &pb.TranscriptSegment{
+		Id:    s.nextID,
+		Start: secondsToNanos(s.cur[0].Start),
+		End:   secondsToNanos(s.cur[len(s.cur)-1].End),
+		Text:  strings.TrimSpace(strings.Join(parts, " ")),
+	})
+	s.nextID++
+	s.cur = nil
+}
+
+func (s *streamSegmenter) segments() []*pb.TranscriptSegment { return s.segs }
+
 // wordsRequested reports whether the caller asked for word-level timestamps.
 // The OpenAI transcription API gates word timings behind
 // timestamp_granularities[] containing "word" and defaults to segment-level
@@ -361,7 +553,12 @@ func (p *ParakeetCpp) AudioTranscriptionStream(ctx context.Context, opts *pb.Tra
 		return status.Error(codes.Canceled, "transcription cancelled")
 	}
 
-	stream := CppStreamBegin(p.ctxPtr)
+	var stream uintptr
+	if CppStreamBeginLang != nil {
+		stream = CppStreamBeginLang(p.ctxPtr, opts.GetLanguage())
+	} else {
+		stream = CppStreamBegin(p.ctxPtr)
+	}
 	if stream == 0 {
 		// Not a cache-aware streaming model: run a normal offline
 		// transcription and emit it as one delta + a closing final result.
@@ -390,6 +587,14 @@ func (p *ParakeetCpp) AudioTranscriptionStream(ctx context.Context, opts *pb.Tra
 		return err
 	}
 
+	// ABI v4: when the streaming JSON entry points are present, drive them so the
+	// per-utterance segments carry per-word start/end timestamps. Falls through to
+	// the text-only loop below against an older libparakeet.so. Runs under the
+	// engineMu already held above.
+	if CppStreamFeedJSON != nil {
+		return p.streamJSON(ctx, stream, data, duration, results)
+	}
+
 	var (
 		full     strings.Builder
 		segText  strings.Builder
@@ -466,6 +671,71 @@ func (p *ParakeetCpp) AudioTranscriptionStream(ctx context.Context, opts *pb.Tra
 	return nil
 }
 
+// streamJSON drives the ABI v4 streaming JSON entry points: each feed/finalize
+// returns a {text,eou,frame_sec,words} document. The newly-finalized text is
+// emitted as a delta (unchanged streaming contract) while words are accumulated
+// into per-utterance segments (closed on EOU) so the closing FinalResult carries
+// timestamped segments. Runs under engineMu (already held by the caller).
+func (p *ParakeetCpp) streamJSON(ctx context.Context, stream uintptr, data []float32,
+	duration float32, results chan *pb.TranscriptStreamResponse) error {
+	var (
+		full strings.Builder
+		seg  streamSegmenter
+	)
+	// consume frees the malloc'd char* (a 0 return is an error), parses the JSON,
+	// emits the delta, and routes words through the segmenter.
+	consume := func(ret uintptr) error {
+		if ret == 0 {
+			msg := CppLastError(p.ctxPtr)
+			if msg == "" {
+				msg = "unknown error"
+			}
+			return fmt.Errorf("parakeet-cpp: stream feed/finalize failed: %s", msg)
+		}
+		raw := goStringFromCPtr(ret)
+		CppFreeString(ret)
+		var doc streamFeedJSON
+		if err := json.Unmarshal([]byte(raw), &doc); err != nil {
+			return fmt.Errorf("parakeet-cpp: decode stream json: %w", err)
+		}
+		if doc.Text != "" {
+			full.WriteString(doc.Text)
+			results <- &pb.TranscriptStreamResponse{Delta: doc.Text}
+		}
+		seg.add(doc)
+		return nil
+	}
+
+	for off := 0; off < len(data); off += streamChunkSamples {
+		if err := ctx.Err(); err != nil {
+			return status.Error(codes.Canceled, "transcription cancelled")
+		}
+		end := min(off+streamChunkSamples, len(data))
+		chunk := data[off:end]
+		if err := consume(CppStreamFeedJSON(stream, chunk, int32(len(chunk)))); err != nil {
+			return err
+		}
+	}
+	if err := consume(CppStreamFinalizeJSON(stream)); err != nil {
+		return err
+	}
+	seg.flush() // close any trailing utterance that never saw an EOU
+
+	text := strings.TrimSpace(full.String())
+	segments := seg.segments()
+	if len(segments) == 0 && text != "" {
+		segments = append(segments, &pb.TranscriptSegment{Id: 0, Text: text})
+	}
+	results <- &pb.TranscriptStreamResponse{
+		FinalResult: &pb.TranscriptResult{
+			Text:     text,
+			Segments: segments,
+			Duration: duration,
+		},
+	}
+	return nil
+}
+
 // decodeWavMono16k converts any input audio to 16 kHz mono PCM and returns the
 // float samples plus the clip duration in seconds. Mirrors the whisper
 // backend: utils.AudioToWav (ffmpeg) normalises rate/channels, go-audio

backend/go/parakeet-cpp/goparakeetcpp_test.go

@@ -53,6 +53,10 @@ func ensureLibLoaded() {
 		purego.RegisterLibFunc(&CppStreamFeed, lib, "parakeet_capi_stream_feed")
 		purego.RegisterLibFunc(&CppStreamFinalize, lib, "parakeet_capi_stream_finalize")
 		purego.RegisterLibFunc(&CppStreamFree, lib, "parakeet_capi_stream_free")
+		if sym, err := purego.Dlsym(lib, "parakeet_capi_stream_feed_json"); err == nil && sym != 0 {
+			purego.RegisterLibFunc(&CppStreamFeedJSON, lib, "parakeet_capi_stream_feed_json")
+			purego.RegisterLibFunc(&CppStreamFinalizeJSON, lib, "parakeet_capi_stream_finalize_json")
+		}
 		purego.RegisterLibFunc(&CppFreeString, lib, "parakeet_capi_free_string")
 		purego.RegisterLibFunc(&CppLastError, lib, "parakeet_capi_last_error")
 	})
@@ -107,13 +111,22 @@ var _ = Describe("ParakeetCpp", func() {
 			Expect(err).ToNot(HaveOccurred())
 			Expect(strings.TrimSpace(res.Text)).ToNot(BeEmpty(),
 				"expected non-empty transcript for %s", audioPath)
-			Expect(res.Segments).To(HaveLen(1),
-				"synthesises a single whole-clip segment")
-			Expect(res.Segments[0].Text).To(Equal(res.Text),
-				"single segment text must equal the top-level text")
-			// Default (no granularities) is segment-level: no per-word timings.
-			Expect(res.Segments[0].Words).To(BeEmpty(),
-				"word timings are opt-in via timestamp_granularities")
+			// NeMo-faithful segmentation: one or more punctuation-delimited
+			// segments, each with text and a monotonically-advancing time span.
+			Expect(res.Segments).ToNot(BeEmpty(), "expected at least one segment")
+			var prevEnd int64
+			for i, seg := range res.Segments {
+				Expect(strings.TrimSpace(seg.Text)).ToNot(BeEmpty(),
+					"segment %d must have text", i)
+				Expect(seg.End).To(BeNumerically(">=", seg.Start),
+					"segment %d end must not precede its start", i)
+				Expect(seg.Start).To(BeNumerically(">=", prevEnd),
+					"segments must be in time order")
+				prevEnd = seg.End
+				// Default (no granularities) is segment-level: no per-word timings.
+				Expect(seg.Words).To(BeEmpty(),
+					"word timings are opt-in via timestamp_granularities")
+			}
 		})
 
 		It("emits word-level timestamps when granularity=word", func() {
@@ -129,15 +142,28 @@ var _ = Describe("ParakeetCpp", func() {
 				TimestampGranularities: []string{"word"},
 			})
 			Expect(err).ToNot(HaveOccurred())
-			Expect(res.Segments).To(HaveLen(1))
-			seg := res.Segments[0]
-			Expect(seg.Words).ToNot(BeEmpty(),
-				"expected per-word timestamps with granularity=word")
-			// Monotonic, non-negative timings spanning the segment.
-			Expect(seg.Words[0].Start).To(BeNumerically(">=", int64(0)))
-			Expect(seg.End).To(BeNumerically(">=", seg.Start))
-			Expect(seg.Words[len(seg.Words)-1].End).To(Equal(seg.End),
-				"segment end tracks the last word")
+			Expect(res.Segments).ToNot(BeEmpty())
+			// With word granularity every segment carries its own words, and each
+			// segment's span tracks its first/last word; word starts advance
+			// monotonically across the whole transcript.
+			totalWords := 0
+			var prevStart int64 = -1
+			for i, seg := range res.Segments {
+				Expect(seg.Words).ToNot(BeEmpty(),
+					"segment %d must carry per-word timestamps with granularity=word", i)
+				Expect(seg.Start).To(Equal(seg.Words[0].Start),
+					"segment %d start tracks its first word", i)
+				Expect(seg.End).To(Equal(seg.Words[len(seg.Words)-1].End),
+					"segment %d end tracks its last word", i)
+				for _, w := range seg.Words {
+					Expect(w.End).To(BeNumerically(">=", w.Start))
+					Expect(w.Start).To(BeNumerically(">=", prevStart))
+					prevStart = w.Start
+					totalWords++
+				}
+			}
+			Expect(totalWords).To(BeNumerically(">", 0))
+			Expect(res.Segments[0].Words[0].Start).To(BeNumerically(">=", int64(0)))
 		})
 	})
 

backend/go/parakeet-cpp/main.go

@@ -65,6 +65,25 @@ func main() {
 		purego.RegisterLibFunc(&CppTranscribePcmBatchJSON, lib, "parakeet_capi_transcribe_pcm_batch_json")
 	}
 
+	// Per-request language variants (multilingual nemotron). Same probe pattern:
+	// present only in libparakeet.so built with multilingual support, so the
+	// backend still loads against an older library and falls back to the
+	// non-lang batched + streaming entry points (model default / "auto").
+	if sym, err := purego.Dlsym(lib, "parakeet_capi_transcribe_pcm_batch_json_lang"); err == nil && sym != 0 {
+		purego.RegisterLibFunc(&CppTranscribePcmBatchJSONLang, lib, "parakeet_capi_transcribe_pcm_batch_json_lang")
+	}
+	if sym, err := purego.Dlsym(lib, "parakeet_capi_stream_begin_lang"); err == nil && sym != 0 {
+		purego.RegisterLibFunc(&CppStreamBeginLang, lib, "parakeet_capi_stream_begin_lang")
+	}
+
+	// Streaming JSON entry points (ABI v4): surface per-word timestamps on the
+	// streaming path. Same probe pattern; absent in older libparakeet.so, where
+	// the backend falls back to the text-only streaming feed.
+	if sym, err := purego.Dlsym(lib, "parakeet_capi_stream_feed_json"); err == nil && sym != 0 {
+		purego.RegisterLibFunc(&CppStreamFeedJSON, lib, "parakeet_capi_stream_feed_json")
+		purego.RegisterLibFunc(&CppStreamFinalizeJSON, lib, "parakeet_capi_stream_finalize_json")
+	}
+
 	fmt.Fprintf(os.Stderr, "[parakeet-cpp] ABI=%d\n", CppAbiVersion())
 
 	flag.Parse()

backend/go/parakeet-cpp/segments_test.go

@@ -0,0 +1,127 @@
+package main
+
+import (
+	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+func tw(text string, start, end float64) transcriptWord {
+	return transcriptWord{W: text, Start: start, End: end}
+}
+
+var _ = Describe("splitWordsIntoSegments (NeMo get_segment_offsets parity)", func() {
+	seps := []rune{'.', '?', '!'}
+
+	It("splits on sentence-ending punctuation, including the delimiter word", func() {
+		words := []transcriptWord{tw("hello", 0, 0.4), tw("world.", 0.4, 0.8), tw("bye", 1.0, 1.3)}
+		segs := splitWordsIntoSegments(words, seps, 0)
+		Expect(segs).To(HaveLen(2))
+		Expect(segs[0]).To(HaveLen(2))
+		Expect(segs[0][1].W).To(Equal("world."))
+		Expect(segs[1]).To(HaveLen(1))
+		Expect(segs[1][0].W).To(Equal("bye"))
+	})
+
+	It("keeps a single segment with no terminal punctuation and gap off", func() {
+		words := []transcriptWord{tw("a", 0, 0.2), tw("b", 0.2, 0.4), tw("c", 5.0, 5.2)}
+		segs := splitWordsIntoSegments(words, seps, 0)
+		Expect(segs).To(HaveLen(1))
+	})
+
+	It("splits on the gap rule when enabled, the gapped word starting the next segment", func() {
+		words := []transcriptWord{tw("a", 0, 0.2), tw("b", 0.2, 0.4), tw("c", 5.0, 5.2)}
+		segs := splitWordsIntoSegments(words, seps, 1.0) // c is 4.6s after b
+		Expect(segs).To(HaveLen(2))
+		Expect(segs[0]).To(HaveLen(2)) // a b
+		Expect(segs[1]).To(HaveLen(1)) // c
+		Expect(segs[1][0].W).To(Equal("c"))
+	})
+
+	It("checks the gap rule before punctuation (NeMo elif order)", func() {
+		// "b." would terminate, but c is far after it -> gap closes [a b.] at b.
+		words := []transcriptWord{tw("a", 0, 0.2), tw("b.", 0.2, 0.4), tw("c", 9.0, 9.2)}
+		segs := splitWordsIntoSegments(words, seps, 1.0)
+		Expect(segs).To(HaveLen(2))
+		Expect(segs[0]).To(HaveLen(2))
+		Expect(segs[1][0].W).To(Equal("c"))
+	})
+
+	It("still splits on punctuation when the gap rule is enabled but does not fire", func() {
+		words := []transcriptWord{tw("hi.", 0, 0.4), tw("bye", 0.4, 0.8)}
+		segs := splitWordsIntoSegments(words, seps, 5.0) // gap never reached
+		Expect(segs).To(HaveLen(2))
+		Expect(segs[0][0].W).To(Equal("hi."))
+	})
+
+	It("returns nothing for empty input", func() {
+		Expect(splitWordsIntoSegments(nil, seps, 0)).To(BeEmpty())
+	})
+})
+
+var _ = Describe("transcriptResultFromDoc (multi-segment)", func() {
+	doc := transcriptJSON{
+		Text:     "hello world. bye now",
+		FrameSec: 0.08,
+		Words: []transcriptWord{
+			{W: "hello", Start: 0.0, End: 0.4},
+			{W: "world.", Start: 0.4, End: 0.8},
+			{W: "bye", Start: 1.0, End: 1.3},
+			{W: "now", Start: 1.3, End: 1.6},
+		},
+		Tokens: []transcriptToken{{ID: 1, T: 0.1}, {ID: 2, T: 0.5}, {ID: 3, T: 1.1}, {ID: 4, T: 1.4}},
+	}
+
+	It("emits one segment per punctuation-delimited group with start/end", func() {
+		res := transcriptResultFromDoc(doc, &pb.TranscriptRequest{}, 0)
+		Expect(res.Segments).To(HaveLen(2))
+		Expect(res.Segments[0].Text).To(Equal("hello world."))
+		Expect(res.Segments[0].Start).To(Equal(int64(0)))
+		Expect(res.Segments[0].End).To(Equal(secondsToNanos(0.8)))
+		Expect(res.Segments[1].Text).To(Equal("bye now"))
+		Expect(res.Segments[1].Start).To(Equal(secondsToNanos(1.0)))
+		Expect(res.Segments[1].Id).To(Equal(int32(1)))
+	})
+
+	It("assigns tokens to the segment whose time window contains them", func() {
+		res := transcriptResultFromDoc(doc, &pb.TranscriptRequest{}, 0)
+		Expect(res.Segments[0].Tokens).To(Equal([]int32{1, 2}))
+		Expect(res.Segments[1].Tokens).To(Equal([]int32{3, 4}))
+	})
+
+	It("attaches per-segment words only when word granularity requested", func() {
+		plain := transcriptResultFromDoc(doc, &pb.TranscriptRequest{}, 0)
+		Expect(plain.Segments[0].Words).To(BeEmpty())
+		withWords := transcriptResultFromDoc(doc, &pb.TranscriptRequest{TimestampGranularities: []string{"word"}}, 0)
+		Expect(withWords.Segments[0].Words).To(HaveLen(2))
+	})
+
+	It("falls back to a single text segment when there are no words", func() {
+		res := transcriptResultFromDoc(transcriptJSON{Text: "hi"}, &pb.TranscriptRequest{}, 0)
+		Expect(res.Segments).To(HaveLen(1))
+		Expect(res.Segments[0].Text).To(Equal("hi"))
+	})
+})
+
+var _ = Describe("streaming segment assembly", func() {
+	It("closes a segment with start/end from its words on EOU", func() {
+		acc := &streamSegmenter{}
+		acc.add(streamFeedJSON{Text: "hello world", Eou: 1, Words: []transcriptWord{
+			{W: "hello", Start: 0.0, End: 0.4}, {W: "world", Start: 0.4, End: 0.9},
+		}})
+		segs := acc.segments()
+		Expect(segs).To(HaveLen(1))
+		Expect(segs[0].Text).To(Equal("hello world"))
+		Expect(segs[0].Start).To(Equal(int64(0)))
+		Expect(segs[0].End).To(Equal(secondsToNanos(0.9)))
+	})
+
+	It("buffers words across feeds until EOU", func() {
+		acc := &streamSegmenter{}
+		acc.add(streamFeedJSON{Text: "hi", Eou: 0, Words: []transcriptWord{{W: "hi", Start: 0, End: 0.3}}})
+		Expect(acc.segments()).To(BeEmpty())
+		acc.add(streamFeedJSON{Text: "there", Eou: 1, Words: []transcriptWord{{W: "there", Start: 0.3, End: 0.7}}})
+		Expect(acc.segments()).To(HaveLen(1))
+		Expect(acc.segments()[0].Text).To(Equal("hi there"))
+	})
+})

backend/go/stablediffusion-ggml/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # stablediffusion.cpp (ggml)
 STABLEDIFFUSION_GGML_REPO?=https://github.com/leejet/stable-diffusion.cpp
-STABLEDIFFUSION_GGML_VERSION?=1f9ee88e09c258053fa59d5e05e23dfb10fa0b13
+STABLEDIFFUSION_GGML_VERSION?=19bdfe22d255d5b4dff39d449318b9bc5ea2317f
 
 CMAKE_ARGS+=-DGGML_MAX_NAME=128
 

backend/go/stablediffusion-ggml/cpp/gosd.cpp

@@ -386,6 +386,7 @@ int load_model(const char *model, char *model_path, char* options[], int threads
     const char *llm_vision_path = "";
     const char *diffusion_model_path = stableDiffusionModel;
     const char *high_noise_diffusion_model_path = "";
+    const char *uncond_diffusion_model_path = "";
     const char *taesd_path  = "";
     const char *control_net_path = "";
     const char *embedding_dir = "";
@@ -472,6 +473,7 @@ int load_model(const char *model, char *model_path, char* options[], int threads
         if (!strcmp(optname, "llm_vision_path")) llm_vision_path = strdup(optval);
         if (!strcmp(optname, "diffusion_model_path")) diffusion_model_path = strdup(optval);
         if (!strcmp(optname, "high_noise_diffusion_model_path")) high_noise_diffusion_model_path = strdup(optval);
+        if (!strcmp(optname, "uncond_diffusion_model_path")) uncond_diffusion_model_path = strdup(optval);
         if (!strcmp(optname, "taesd_path")) taesd_path = strdup(optval);
         if (!strcmp(optname, "control_net_path")) control_net_path = strdup(optval);
         if (!strcmp(optname, "embedding_dir")) {
@@ -571,6 +573,7 @@ int load_model(const char *model, char *model_path, char* options[], int threads
     ctx_params.llm_vision_path = llm_vision_path;
     ctx_params.diffusion_model_path = diffusion_model_path;
     ctx_params.high_noise_diffusion_model_path = high_noise_diffusion_model_path;
+    ctx_params.uncond_diffusion_model_path = uncond_diffusion_model_path;
     ctx_params.vae_path = vae_path;
     ctx_params.audio_vae_path = audio_vae_path;
     ctx_params.embeddings_connectors_path = embeddings_connectors_path;

backend/go/whisper/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # whisper.cpp version
 WHISPER_REPO?=https://github.com/ggml-org/whisper.cpp
-WHISPER_CPP_VERSION?=99613cb720b65036237d44b52f753b51f75c2797
+WHISPER_CPP_VERSION?=df7638d8229a243af8a4b5a8ae557e0d74e0a0ae
 SO_TARGET?=libgowhisper.so
 
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF

backend/python/rerankers/requirements-cpu.txt

@@ -1,4 +1,4 @@
 transformers
 accelerate
-torch==2.7.1+xpu
+torch==2.7.1
 rerankers[transformers]

backend/python/rerankers/requirements-cublas12.txt

@@ -1,4 +1,4 @@
 transformers
 accelerate
-torch==2.7.1+xpu
+torch==2.7.1
 rerankers[transformers]

backend/python/rerankers/requirements-cublas13.txt

@@ -1,5 +1,5 @@
 --extra-index-url https://download.pytorch.org/whl/cu130
 transformers
 accelerate
-torch==2.7.1+xpu
+torch==2.9.1
 rerankers[transformers]

backend/python/rerankers/requirements-hipblas.txt

@@ -1,5 +1,5 @@
 --extra-index-url https://download.pytorch.org/whl/rocm7.0
 transformers
 accelerate
-torch==2.7.1+xpu
+torch==2.10.0+rocm7.0
 rerankers[transformers]

backend/python/rerankers/requirements-mps.txt

@@ -1,4 +1,4 @@
-torch==2.7.1+xpu
+torch==2.7.1
 transformers
 accelerate
 rerankers[transformers]

core/application/startup.go

@@ -23,9 +23,9 @@ import (
 	"github.com/mudler/LocalAI/core/services/routing/pii"
 	"github.com/mudler/LocalAI/core/services/routing/router"
 	"github.com/mudler/LocalAI/core/services/storage"
-	"github.com/mudler/LocalAI/pkg/signals"
 	coreStartup "github.com/mudler/LocalAI/core/startup"
 	"github.com/mudler/LocalAI/internal"
+	"github.com/mudler/LocalAI/pkg/signals"
 	"github.com/mudler/LocalAI/pkg/vram"
 
 	"github.com/mudler/LocalAI/pkg/model"
@@ -308,10 +308,31 @@ func New(opts ...config.AppOption) (*Application, error) {
 			application.galleryService.SetNATSClient(distSvc.Nats)
 			if distSvc.DistStores != nil && distSvc.DistStores.Gallery != nil {
 				// Clean up stale in-progress operations from previous crashed instances
-				if err := distSvc.DistStores.Gallery.CleanStale(30 * time.Minute); err != nil {
+				if _, err := distSvc.DistStores.Gallery.CleanStale(30 * time.Minute); err != nil {
 					xlog.Warn("Failed to clean stale gallery operations", "error", err)
 				}
 				application.galleryService.SetGalleryStore(distSvc.DistStores.Gallery)
+
+				// Reap stale ops periodically, not just at boot: an op orphaned by
+				// a replica that died mid-install (its foreground handler goroutine
+				// gone) would otherwise linger "processing" in the UI until the next
+				// restart. 30m matches the install/upgrade ceiling so a genuinely
+				// slow op is never reaped out from under itself.
+				gsvc := application.galleryService
+				go func() {
+					ticker := time.NewTicker(15 * time.Minute)
+					defer ticker.Stop()
+					for {
+						select {
+						case <-options.Context.Done():
+							return
+						case <-ticker.C:
+							if _, err := gsvc.ReapStaleOperations(30 * time.Minute); err != nil {
+								xlog.Warn("Failed to reap stale gallery operations", "error", err)
+							}
+						}
+					}
+				}()
 			}
 			// Hydrate from the store first so the wildcard subscriber finds an
 			// already-populated statuses map for any operations still in flight

core/application/upgrade_checker.go

@@ -214,7 +214,9 @@ func (uc *UpgradeChecker) runCheck(ctx context.Context) {
 				"from", info.InstalledVersion, "to", info.AvailableVersion)
 			var err error
 			if bm != nil {
-				err = bm.UpgradeBackend(ctx, name, nil)
+				// Background auto-upgrade: no live admin watching a progress bar,
+				// so opID is empty and the distributed path skips progress streaming.
+				err = bm.UpgradeBackend(ctx, "", name, nil)
 			} else {
 				err = gallery.UpgradeBackend(ctx, uc.systemState, uc.modelLoader,
 					uc.galleries, name, nil, uc.appConfig.RequireBackendIntegrity)

core/cli/chat/chat.go

@@ -0,0 +1,30 @@
+package chat
+
+import (
+	"context"
+	"io"
+	"strings"
+)
+
+type Options struct {
+	Model   string
+	BaseURL string
+	APIKey  string
+	In      io.Reader
+	Out     io.Writer
+}
+
+func Run(ctx context.Context, opts Options) error {
+	if opts.In == nil {
+		opts.In = strings.NewReader("")
+	}
+	if opts.Out == nil {
+		opts.Out = io.Discard
+	}
+
+	session, err := newChatSession(ctx, newLocalAIChatClient(opts.BaseURL, opts.APIKey), opts.Model)
+	if err != nil {
+		return err
+	}
+	return runTerminalChat(ctx, session, opts.In, opts.Out)
+}

core/cli/chat/chat_suite_test.go

@@ -0,0 +1,13 @@
+package chat
+
+import (
+	"testing"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+func TestChat(t *testing.T) {
+	RegisterFailHandler(Fail)
+	RunSpecs(t, "Chat Suite")
+}

core/cli/chat/chat_test.go

@@ -0,0 +1,172 @@
+package chat
+
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("Run chat", func() {
+	It("streams a single chat response", func() {
+		var capturedModel string
+		var capturedAuth string
+		server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+			if r.URL.Path == "/v1/models" {
+				w.Header().Set("Content-Type", "application/json")
+				writeResponse(w, `{"object":"list","data":[{"id":"test-model","object":"model"}]}`)
+				return
+			}
+
+			Expect(r.URL.Path).To(Equal("/v1/chat/completions"))
+			capturedAuth = r.Header.Get("Authorization")
+
+			var body struct {
+				Model    string `json:"model"`
+				Messages []struct {
+					Role    string `json:"role"`
+					Content string `json:"content"`
+				} `json:"messages"`
+			}
+			Expect(json.NewDecoder(r.Body).Decode(&body)).To(Succeed())
+			capturedModel = body.Model
+			Expect(body.Messages).To(HaveLen(1))
+			Expect(body.Messages[0].Role).To(Equal("user"))
+			Expect(body.Messages[0].Content).To(Equal("hello"))
+
+			w.Header().Set("Content-Type", "text/event-stream")
+			writeResponse(w, "data: {\"choices\":[{\"index\":0,\"delta\":{\"content\":\"hi\"}}]}\n\n")
+			writeResponse(w, "data: {\"choices\":[{\"index\":0,\"delta\":{\"content\":\"!\"}}]}\n\n")
+			writeResponse(w, "data: [DONE]\n\n")
+		}))
+		defer server.Close()
+
+		var out bytes.Buffer
+		err := Run(GinkgoT().Context(), Options{
+			Model:   "test-model",
+			BaseURL: server.URL + "/v1",
+			APIKey:  "secret",
+			In:      strings.NewReader("hello\n/exit\n"),
+			Out:     &out,
+		})
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(capturedModel).To(Equal("test-model"))
+		Expect(capturedAuth).To(Equal("Bearer secret"))
+		Expect(out.String()).To(ContainSubstring("assistant: hi!"))
+		Expect(out.String()).To(ContainSubstring("bye"))
+	})
+
+	It("auto-selects the only available model", func() {
+		server := chatTestServer([]string{"solo"}, nil)
+		defer server.Close()
+
+		var out bytes.Buffer
+		err := Run(GinkgoT().Context(), Options{
+			BaseURL: server.URL + "/v1",
+			In:      strings.NewReader("/exit\n"),
+			Out:     &out,
+		})
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(out.String()).To(ContainSubstring("LocalAI chat (solo)"))
+	})
+
+	It("returns an actionable error when no models are installed", func() {
+		server := chatTestServer(nil, nil)
+		defer server.Close()
+
+		err := Run(GinkgoT().Context(), Options{
+			BaseURL: server.URL + "/v1",
+			In:      strings.NewReader(""),
+		})
+
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("no chat models are installed"))
+		Expect(err.Error()).To(ContainSubstring("local-ai models install <model>"))
+	})
+
+	It("returns an actionable error when multiple models are available without a selection", func() {
+		server := chatTestServer([]string{"alpha", "beta"}, nil)
+		defer server.Close()
+
+		err := Run(GinkgoT().Context(), Options{
+			BaseURL: server.URL + "/v1",
+			In:      strings.NewReader(""),
+		})
+
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("multiple models are available"))
+		Expect(err.Error()).To(ContainSubstring("--model"))
+		Expect(err.Error()).To(ContainSubstring("alpha"))
+		Expect(err.Error()).To(ContainSubstring("beta"))
+	})
+
+	It("lists and switches models inside the chat", func() {
+		requestedModels := []string{}
+		server := chatTestServer([]string{"alpha", "beta"}, func(model string) {
+			requestedModels = append(requestedModels, model)
+		})
+		defer server.Close()
+
+		var out bytes.Buffer
+		err := Run(GinkgoT().Context(), Options{
+			Model:   "alpha",
+			BaseURL: server.URL + "/v1",
+			In:      strings.NewReader("/models\n/model beta\nhello\n/exit\n"),
+			Out:     &out,
+		})
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(out.String()).To(ContainSubstring("* alpha"))
+		Expect(out.String()).To(ContainSubstring("  beta"))
+		Expect(out.String()).To(ContainSubstring("switched to beta; conversation cleared"))
+		Expect(requestedModels).To(Equal([]string{"beta"}))
+	})
+})
+
+func chatTestServer(models []string, onChat func(model string)) *httptest.Server {
+	return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		switch r.URL.Path {
+		case "/v1/models":
+			w.Header().Set("Content-Type", "application/json")
+			writeResponse(w, `{"object":"list","data":[`)
+			for i, model := range models {
+				if i > 0 {
+					writeResponse(w, ",")
+				}
+				writeResponsef(w, `{"id":%q,"object":"model"}`, model)
+			}
+			writeResponse(w, `]}`)
+		case "/v1/chat/completions":
+			var body struct {
+				Model string `json:"model"`
+			}
+			Expect(json.NewDecoder(r.Body).Decode(&body)).To(Succeed())
+			if onChat != nil {
+				onChat(body.Model)
+			}
+			w.Header().Set("Content-Type", "text/event-stream")
+			writeResponse(w, "data: {\"choices\":[{\"index\":0,\"delta\":{\"content\":\"ok\"}}]}\n\n")
+			writeResponse(w, "data: [DONE]\n\n")
+		default:
+			w.WriteHeader(http.StatusNotFound)
+		}
+	}))
+}
+
+func writeResponse(w io.Writer, text string) {
+	_, err := fmt.Fprint(w, text)
+	Expect(err).ToNot(HaveOccurred())
+}
+
+func writeResponsef(w io.Writer, format string, args ...any) {
+	_, err := fmt.Fprintf(w, format, args...)
+	Expect(err).ToNot(HaveOccurred())
+}

core/cli/chat/client.go

@@ -0,0 +1,114 @@
+package chat
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"sort"
+	"strings"
+
+	openai "github.com/sashabaranov/go-openai"
+)
+
+type chatClient interface {
+	ListModels(ctx context.Context) ([]string, error)
+	StreamChat(ctx context.Context, model string, messages []chatMessage, out io.Writer) (string, error)
+}
+
+type localAIChatClient struct {
+	client *openai.Client
+}
+
+func newLocalAIChatClient(baseURL string, apiKey string) *localAIChatClient {
+	cfg := openai.DefaultConfig(apiKey)
+	cfg.BaseURL = baseURL
+	return &localAIChatClient{client: openai.NewClientWithConfig(cfg)}
+}
+
+func (c *localAIChatClient) ListModels(ctx context.Context) ([]string, error) {
+	resp, err := c.client.ListModels(ctx)
+	if err != nil {
+		return nil, err
+	}
+
+	models := make([]string, 0, len(resp.Models))
+	for _, model := range resp.Models {
+		if model.ID != "" {
+			models = append(models, model.ID)
+		}
+	}
+	sort.Strings(models)
+	return models, nil
+}
+
+func (c *localAIChatClient) StreamChat(ctx context.Context, model string, messages []chatMessage, out io.Writer) (string, error) {
+	stream, err := c.client.CreateChatCompletionStream(ctx, openai.ChatCompletionRequest{
+		Model:    model,
+		Messages: openAIChatMessages(messages),
+	})
+	if err != nil {
+		return "", friendlyChatError(err, model)
+	}
+	defer func() {
+		_ = stream.Close()
+	}()
+
+	var answer strings.Builder
+	for {
+		resp, err := stream.Recv()
+		if errors.Is(err, io.EOF) {
+			break
+		}
+		if err != nil {
+			return answer.String(), friendlyChatError(err, model)
+		}
+		if len(resp.Choices) == 0 {
+			continue
+		}
+
+		token := resp.Choices[0].Delta.Content
+		if token == "" {
+			continue
+		}
+		answer.WriteString(token)
+		if _, err := fmt.Fprint(out, token); err != nil {
+			return answer.String(), err
+		}
+	}
+
+	return answer.String(), nil
+}
+
+func openAIChatMessages(messages []chatMessage) []openai.ChatCompletionMessage {
+	converted := make([]openai.ChatCompletionMessage, len(messages))
+	for i, message := range messages {
+		converted[i] = openai.ChatCompletionMessage{
+			Role:    message.Role,
+			Content: message.Content,
+		}
+	}
+	return converted
+}
+
+func friendlyChatError(err error, model string) error {
+	var apiErr *openai.APIError
+	if errors.As(err, &apiErr) {
+		switch apiErr.HTTPStatusCode {
+		case 404:
+			return fmt.Errorf("model %q is not available. Run `local-ai models list`, install a model with `local-ai models install <model>`, or switch with `/model <name>`", model)
+		case 403:
+			return fmt.Errorf("model %q is disabled. Enable it from LocalAI settings or choose another model with `/model <name>`", model)
+		}
+		if apiErr.Message != "" {
+			return errors.New(apiErr.Message)
+		}
+	}
+
+	msg := err.Error()
+	if strings.Contains(msg, "model") && strings.Contains(msg, "not found") {
+		return fmt.Errorf("model %q is not available. Run `local-ai models list`, install a model with `local-ai models install <model>`, or switch with `/model <name>`", model)
+	}
+
+	return err
+}

core/cli/chat/models.go

@@ -0,0 +1,17 @@
+package chat
+
+import "strings"
+
+func formatChatModelList(models []string, current string) string {
+	var b strings.Builder
+	for _, model := range models {
+		prefix := "  "
+		if model == current {
+			prefix = "* "
+		}
+		b.WriteString(prefix)
+		b.WriteString(model)
+		b.WriteByte('\n')
+	}
+	return b.String()
+}

core/cli/chat/session.go

@@ -0,0 +1,120 @@
+package chat
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"strings"
+)
+
+const (
+	chatRoleUser      = "user"
+	chatRoleAssistant = "assistant"
+)
+
+type chatMessage struct {
+	Role    string
+	Content string
+}
+
+type chatSession struct {
+	client   chatClient
+	model    string
+	models   []string
+	messages []chatMessage
+}
+
+func newChatSession(ctx context.Context, client chatClient, requestedModel string) (*chatSession, error) {
+	models, err := client.ListModels(ctx)
+	if err != nil {
+		return nil, fmt.Errorf("list models: %w", err)
+	}
+
+	model, err := resolveChatModel(requestedModel, models)
+	if err != nil {
+		return nil, err
+	}
+
+	return &chatSession{
+		client: client,
+		model:  model,
+		models: models,
+	}, nil
+}
+
+func (s *chatSession) CurrentModel() string {
+	return s.model
+}
+
+func (s *chatSession) Models() []string {
+	models := make([]string, len(s.models))
+	copy(models, s.models)
+	return models
+}
+
+func (s *chatSession) Clear() {
+	s.messages = nil
+}
+
+func (s *chatSession) SwitchModel(model string) error {
+	if !modelExists(s.models, model) {
+		return fmt.Errorf("model %q is not available. Use /models to see installed models", model)
+	}
+	s.model = model
+	s.Clear()
+	return nil
+}
+
+func (s *chatSession) Send(ctx context.Context, prompt string, out io.Writer) error {
+	s.messages = append(s.messages, chatMessage{
+		Role:    chatRoleUser,
+		Content: prompt,
+	})
+
+	answer, err := s.client.StreamChat(ctx, s.model, s.messages, out)
+	if err != nil {
+		return err
+	}
+
+	s.messages = append(s.messages, chatMessage{
+		Role:    chatRoleAssistant,
+		Content: answer,
+	})
+	return nil
+}
+
+func resolveChatModel(requested string, models []string) (string, error) {
+	switch {
+	case requested == "" && len(models) == 0:
+		return "", errors.New(`no chat models are installed.
+
+Install a model first, for example:
+  local-ai models list
+  local-ai models install <model>
+  local-ai run
+
+Then start a chat session:
+  local-ai chat --model <model>`)
+	case requested == "" && len(models) == 1:
+		return models[0], nil
+	case requested == "" && len(models) > 1:
+		var b strings.Builder
+		b.WriteString("multiple models are available; choose one with --model:\n")
+		b.WriteString(formatChatModelList(models, ""))
+		return "", errors.New(b.String())
+	case !modelExists(models, requested):
+		return "", fmt.Errorf("model %q is not available. Use `local-ai models list` and `local-ai models install <model>`, or pass an installed model with --model", requested)
+	default:
+		return requested, nil
+	}
+}
+
+func modelExists(models []string, name string) bool {
+	for _, model := range models {
+		if model == name {
+			return true
+		}
+	}
+	return false
+}

core/cli/chat/session_test.go

@@ -0,0 +1,56 @@
+package chat
+
+import (
+	"context"
+	"io"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("Chat session", func() {
+	It("keeps model switching and message history out of the terminal adapter", func() {
+		client := &fakeChatClient{
+			models: []string{"alpha", "beta"},
+			answer: "pong",
+		}
+
+		session, err := newChatSession(context.Background(), client, "alpha")
+		Expect(err).ToNot(HaveOccurred())
+		Expect(session.CurrentModel()).To(Equal("alpha"))
+
+		Expect(session.SwitchModel("beta")).To(Succeed())
+		Expect(session.CurrentModel()).To(Equal("beta"))
+		Expect(session.Send(context.Background(), "ping", io.Discard)).To(Succeed())
+
+		Expect(client.requests).To(HaveLen(1))
+		Expect(client.requests[0].model).To(Equal("beta"))
+		Expect(client.requests[0].messages).To(HaveLen(1))
+		Expect(client.requests[0].messages[0].Content).To(Equal("ping"))
+	})
+})
+
+type fakeChatClient struct {
+	models   []string
+	answer   string
+	requests []fakeChatRequest
+}
+
+type fakeChatRequest struct {
+	model    string
+	messages []chatMessage
+}
+
+func (c *fakeChatClient) ListModels(context.Context) ([]string, error) {
+	return c.models, nil
+}
+
+func (c *fakeChatClient) StreamChat(_ context.Context, model string, messages []chatMessage, out io.Writer) (string, error) {
+	copied := make([]chatMessage, len(messages))
+	copy(copied, messages)
+	c.requests = append(c.requests, fakeChatRequest{model: model, messages: copied})
+	if _, err := io.WriteString(out, c.answer); err != nil {
+		return "", err
+	}
+	return c.answer, nil
+}

core/cli/chat/terminal.go

@@ -0,0 +1,93 @@
+package chat
+
+import (
+	"bufio"
+	"context"
+	"fmt"
+	"io"
+	"strings"
+)
+
+func runTerminalChat(ctx context.Context, session *chatSession, in io.Reader, out io.Writer) error {
+	scanner := bufio.NewScanner(in)
+	scanner.Buffer(make([]byte, 0, 64*1024), 4*1024*1024)
+
+	if err := writeChat(out, "LocalAI chat (%s)\n", session.CurrentModel()); err != nil {
+		return err
+	}
+	if err := writeChat(out, "Type /exit to quit, /clear to reset the conversation, /models to list models.\n"); err != nil {
+		return err
+	}
+
+	for {
+		if err := writeChat(out, "\n> "); err != nil {
+			return err
+		}
+		if !scanner.Scan() {
+			break
+		}
+
+		prompt := strings.TrimSpace(scanner.Text())
+		switch prompt {
+		case "":
+			continue
+		case "/bye", "/exit", "/quit":
+			return writeChat(out, "bye\n")
+		case "/clear":
+			session.Clear()
+			if err := writeChat(out, "conversation cleared\n"); err != nil {
+				return err
+			}
+			continue
+		case "/models":
+			if err := printChatModels(out, session.Models(), session.CurrentModel()); err != nil {
+				return err
+			}
+			continue
+		}
+
+		if nextModel, ok := strings.CutPrefix(prompt, "/model "); ok {
+			nextModel = strings.TrimSpace(nextModel)
+			if nextModel == "" {
+				if err := writeChat(out, "usage: /model <name>\n"); err != nil {
+					return err
+				}
+				continue
+			}
+			if err := session.SwitchModel(nextModel); err != nil {
+				if writeErr := writeChat(out, "%s\n", err); writeErr != nil {
+					return writeErr
+				}
+				continue
+			}
+			if err := writeChat(out, "switched to %s; conversation cleared\n", session.CurrentModel()); err != nil {
+				return err
+			}
+			continue
+		}
+
+		if err := writeChat(out, "assistant: "); err != nil {
+			return err
+		}
+		if err := session.Send(ctx, prompt, out); err != nil {
+			return err
+		}
+		if err := writeChat(out, "\n"); err != nil {
+			return err
+		}
+	}
+
+	return scanner.Err()
+}
+
+func printChatModels(out io.Writer, models []string, current string) error {
+	if len(models) == 0 {
+		return writeChat(out, "no models installed\n")
+	}
+	return writeChat(out, "%s", formatChatModelList(models, current))
+}
+
+func writeChat(out io.Writer, format string, args ...any) error {
+	_, err := fmt.Fprintf(out, format, args...)
+	return err
+}

core/cli/chat_cmd.go

@@ -0,0 +1,25 @@
+package cli
+
+import (
+	"context"
+	"os"
+
+	chatcli "github.com/mudler/LocalAI/core/cli/chat"
+	cliContext "github.com/mudler/LocalAI/core/cli/context"
+)
+
+type ChatCMD struct {
+	Model    string `short:"m" help:"Model name to use. Defaults to the only model returned by the server when exactly one is available"`
+	Endpoint string `env:"LOCALAI_CHAT_ENDPOINT" default:"http://127.0.0.1:8080" help:"LocalAI server endpoint. The /v1 path is added automatically when omitted"`
+	APIKey   string `env:"LOCALAI_API_KEY,API_KEY" help:"API key to use when the LocalAI server requires authentication"`
+}
+
+func (c *ChatCMD) Run(ctx *cliContext.Context) error {
+	return chatcli.Run(context.Background(), chatcli.Options{
+		Model:   c.Model,
+		BaseURL: chatAPIBaseURL(c.Endpoint),
+		APIKey:  c.APIKey,
+		In:      os.Stdin,
+		Out:     os.Stdout,
+	})
+}

core/cli/chat_cmd_test.go

@@ -0,0 +1,27 @@
+package cli
+
+import (
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("Chat command wiring", func() {
+	Describe("chatAPIBaseURL", func() {
+		It("adds /v1 to a root endpoint", func() {
+			Expect(chatAPIBaseURL("http://127.0.0.1:8080")).To(Equal("http://127.0.0.1:8080/v1"))
+		})
+
+		It("keeps endpoints that already include /v1", func() {
+			Expect(chatAPIBaseURL("http://127.0.0.1:8080/v1")).To(Equal("http://127.0.0.1:8080/v1"))
+			Expect(chatAPIBaseURL("http://127.0.0.1:8080/v1/")).To(Equal("http://127.0.0.1:8080/v1"))
+		})
+
+		It("adds a default http scheme", func() {
+			Expect(chatAPIBaseURL("127.0.0.1:8080")).To(Equal("http://127.0.0.1:8080/v1"))
+		})
+
+		It("preserves non-root paths before /v1", func() {
+			Expect(chatAPIBaseURL("http://127.0.0.1:8080/localai")).To(Equal("http://127.0.0.1:8080/localai/v1"))
+		})
+	})
+})

core/cli/chat_endpoint.go

@@ -0,0 +1,29 @@
+package cli
+
+import (
+	"net/url"
+	"strings"
+)
+
+func chatAPIBaseURL(endpoint string) string {
+	if !strings.Contains(endpoint, "://") {
+		endpoint = "http://" + endpoint
+	}
+
+	u, err := url.Parse(endpoint)
+	if err != nil {
+		return strings.TrimRight(endpoint, "/") + "/v1"
+	}
+
+	path := strings.TrimRight(u.Path, "/")
+	if path == "" {
+		u.Path = "/v1"
+	} else if path != "/v1" && !strings.HasSuffix(path, "/v1") {
+		u.Path = path + "/v1"
+	} else {
+		u.Path = path
+	}
+	u.RawQuery = ""
+	u.Fragment = ""
+	return u.String()
+}

core/cli/cli.go

@@ -9,6 +9,7 @@ var CLI struct {
 	cliContext.Context `embed:""`
 
 	Run             RunCMD             `cmd:"" help:"Run LocalAI, this the default command if no other command is specified. Run 'local-ai run --help' for more information" default:"withargs"`
+	Chat            ChatCMD            `cmd:"" help:"Open an interactive chat session against a running LocalAI server"`
 	Federated       FederatedCLI       `cmd:"" help:"Run LocalAI in federated mode"`
 	Models          ModelsCMD          `cmd:"" help:"Manage LocalAI models and definitions"`
 	Backends        BackendsCMD        `cmd:"" help:"Manage LocalAI backends and definitions"`

core/cli/run.go

@@ -30,6 +30,8 @@ type RunCMD struct {
 	ModelArgs []string `arg:"" optional:"" name:"models" help:"Model configuration URLs to load"`
 
 	ExternalBackends             []string      `env:"LOCALAI_EXTERNAL_BACKENDS,EXTERNAL_BACKENDS" help:"A list of external backends to load from gallery on boot" group:"backends"`
+	WebRTCNAT1To1IPs             []string      `env:"LOCALAI_WEBRTC_NAT_1TO1_IPS,WEBRTC_NAT_1TO1_IPS" help:"IPs advertised as the host ICE candidates for /v1/realtime WebRTC instead of every local interface. Set to the reachable host/LAN IP when running under Docker host networking or NAT, where pion otherwise offers unreachable bridge addresses and the connection drops after ICE consent checks fail." group:"api"`
+	WebRTCICEInterfaces          []string      `env:"LOCALAI_WEBRTC_ICE_INTERFACES,WEBRTC_ICE_INTERFACES" help:"Restrict /v1/realtime WebRTC ICE candidate gathering to these network interfaces (e.g. eth0), filtering out docker0/veth noise." group:"api"`
 	BackendsPath                 string        `env:"LOCALAI_BACKENDS_PATH,BACKENDS_PATH" type:"path" default:"${basepath}/backends" help:"Path containing backends used for inferencing" group:"backends"`
 	BackendsSystemPath           string        `env:"LOCALAI_BACKENDS_SYSTEM_PATH,BACKEND_SYSTEM_PATH" type:"path" default:"/var/lib/local-ai/backends" help:"Path containing system backends used for inferencing" group:"backends"`
 	ModelsPath                   string        `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
@@ -225,6 +227,8 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
 		config.WithApiKeys(r.APIKeys),
 		config.WithModelsURL(append(r.Models, r.ModelArgs...)...),
 		config.WithExternalBackends(r.ExternalBackends...),
+		config.WithWebRTCNAT1To1IPs(r.WebRTCNAT1To1IPs...),
+		config.WithWebRTCICEInterfaces(r.WebRTCICEInterfaces...),
 		config.WithOpaqueErrors(r.OpaqueErrors),
 		config.WithEnforcedPredownloadScans(!r.DisablePredownloadScan),
 		config.WithSubtleKeyComparison(r.UseSubtleKeyComparison),
@@ -652,12 +656,12 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
 // waitForServerReady polls the given address until the HTTP server is
 // accepting connections or the context is cancelled.
 func waitForServerReady(address string, ctx context.Context) {
-	// Ensure the address has a host component for dialing.
-	// Echo accepts ":8080" but net.Dial needs a resolvable host.
 	host, port, err := net.SplitHostPort(address)
 	if err == nil && host == "" {
 		address = "127.0.0.1:" + port
 	}
+	ticker := time.NewTicker(250 * time.Millisecond)
+	defer ticker.Stop()
 
 	for {
 		select {
@@ -665,11 +669,17 @@ func waitForServerReady(address string, ctx context.Context) {
 			return
 		default:
 		}
+
 		conn, err := net.DialTimeout("tcp", address, 500*time.Millisecond)
 		if err == nil {
 			conn.Close()
 			return
 		}
-		time.Sleep(250 * time.Millisecond)
+
+		select {
+		case <-ctx.Done():
+			return
+		case <-ticker.C:
+		}
 	}
 }

core/config/application_config.go

@@ -12,10 +12,19 @@ import (
 )
 
 type ApplicationConfig struct {
-	Context                             context.Context
-	ConfigFile                          string
-	SystemState                         *system.SystemState
-	ExternalBackends                    []string
+	Context          context.Context
+	ConfigFile       string
+	SystemState      *system.SystemState
+	ExternalBackends []string
+
+	// WebRTCNAT1To1IPs, when set, are advertised as the host ICE candidates for
+	// /v1/realtime WebRTC instead of every local interface address. Needed when
+	// the routable address differs from what pion gathers — e.g. Docker host
+	// networking (where pion also offers unreachable bridge IPs) or NAT.
+	WebRTCNAT1To1IPs []string
+	// WebRTCICEInterfaces, when set, restricts ICE candidate gathering to these
+	// network interfaces (e.g. eth0), filtering out docker0/veth noise.
+	WebRTCICEInterfaces                 []string
 	UploadLimitMB, Threads, ContextSize int
 	F16                                 bool
 	Debug                               bool
@@ -81,7 +90,6 @@ type ApplicationConfig struct {
 	// file is mode 0600.
 	MITMCADir string
 
-
 	// PIIPatternOverrides applies persisted per-id deltas (action,
 	// disabled) to the live redactor at startup. Loaded from
 	// runtime_settings.json and applied right after pii.NewRedactor.
@@ -116,11 +124,11 @@ type ApplicationConfig struct {
 	// --require-backend-integrity / LOCALAI_REQUIRE_BACKEND_INTEGRITY.
 	RequireBackendIntegrity bool
 
-	SingleBackend           bool // Deprecated: use MaxActiveBackends = 1 instead
-	MaxActiveBackends       int  // Maximum number of active backends (0 = unlimited, 1 = single backend mode)
-	WatchDogIdle bool
-	WatchDogBusy bool
-	WatchDog     bool
+	SingleBackend     bool // Deprecated: use MaxActiveBackends = 1 instead
+	MaxActiveBackends int  // Maximum number of active backends (0 = unlimited, 1 = single backend mode)
+	WatchDogIdle      bool
+	WatchDogBusy      bool
+	WatchDog          bool
 
 	// Memory Reclaimer settings (works with GPU if available, otherwise RAM)
 	MemoryReclaimerEnabled   bool    // Enable memory threshold monitoring
@@ -311,6 +319,18 @@ func WithExternalBackends(backends ...string) AppOption {
 	}
 }
 
+func WithWebRTCNAT1To1IPs(ips ...string) AppOption {
+	return func(o *ApplicationConfig) {
+		o.WebRTCNAT1To1IPs = ips
+	}
+}
+
+func WithWebRTCICEInterfaces(interfaces ...string) AppOption {
+	return func(o *ApplicationConfig) {
+		o.WebRTCICEInterfaces = interfaces
+	}
+}
+
 func WithMachineTag(tag string) AppOption {
 	return func(o *ApplicationConfig) {
 		o.MachineTag = tag
@@ -702,7 +722,6 @@ func WithMITMCADir(dir string) AppOption {
 	}
 }
 
-
 func WithDynamicConfigDir(dynamicConfigsDir string) AppOption {
 	return func(o *ApplicationConfig) {
 		o.DynamicConfigsDir = dynamicConfigsDir

core/config/hooks_llamacpp.go

@@ -39,7 +39,21 @@ func llamaCppDefaults(cfg *ModelConfig, modelPath string) {
 		}
 	}()
 
-	f, err := gguf.ParseGGUFFile(guessPath)
+	// Startup parses every model's GGUF header to guess defaults. We only need
+	// scalar metadata (architecture, head/ff counts, chat_template, token IDs,
+	// MTP head) plus array *lengths* — never the array *contents*. Two options
+	// keep this cheap, which matters when many models live on slow storage such
+	// as a Docker volume (see https://github.com/mudler/LocalAI/issues/9790):
+	//
+	//   - SkipLargeMetadata: seek past large array-valued metadata (the tokenizer
+	//     vocab: tokenizer.ggml.tokens/scores/merges, often >100k entries) instead
+	//     of reading and allocating every element. Lengths stay populated.
+	//   - UseMMap: read the header via a memory map so faulting in a few pages
+	//     replaces hundreds of thousands of tiny read() syscalls (measured ~524k
+	//     -> 8 for a 256k-token vocab), the dominant cost on slow filesystems.
+	//
+	// The mapping is released when ParseGGUFFile returns.
+	f, err := gguf.ParseGGUFFile(guessPath, gguf.UseMMap(), gguf.SkipLargeMetadata())
 	if err == nil {
 		guessGGUFFromFile(cfg, f, 0)
 	}

core/config/hooks_test.go

@@ -1,13 +1,76 @@
 package config_test
 
 import (
+	"bytes"
+	"encoding/binary"
+	"os"
+	"path/filepath"
+
 	. "github.com/mudler/LocalAI/core/config"
 	"github.com/mudler/LocalAI/core/schema"
 
+	gguf "github.com/gpustack/gguf-parser-go"
 	. "github.com/onsi/ginkgo/v2"
 	. "github.com/onsi/gomega"
 )
 
+// GGUF metadata value type tags (see github.com/gpustack/gguf-parser-go).
+const (
+	ggufTypeUint32 uint32 = 4
+	ggufTypeString uint32 = 8
+	ggufTypeArray  uint32 = 9
+)
+
+// writeTestGGUF emits a minimal but valid little-endian GGUF v3 header carrying
+// the scalar metadata the llama-cpp hook guesses from plus a large string vocab
+// array (tokenizer.ggml.tokens). The big array is exactly what SkipLargeMetadata
+// + UseMMap are expected to avoid reading element-by-element, so it must survive a
+// round-trip through the real hook without corrupting the guessed defaults.
+func writeTestGGUF(path, chatTemplate string, vocab int) error {
+	wStr := func(b *bytes.Buffer, s string) {
+		binary.Write(b, binary.LittleEndian, uint64(len(s)))
+		b.WriteString(s)
+	}
+	kvStr := func(b *bytes.Buffer, k, v string) {
+		wStr(b, k)
+		binary.Write(b, binary.LittleEndian, ggufTypeString)
+		wStr(b, v)
+	}
+	kvU32 := func(b *bytes.Buffer, k string, v uint32) {
+		wStr(b, k)
+		binary.Write(b, binary.LittleEndian, ggufTypeUint32)
+		binary.Write(b, binary.LittleEndian, v)
+	}
+
+	var meta bytes.Buffer
+	kvStr(&meta, "general.architecture", "llama")
+	kvStr(&meta, "general.name", "ReproModel")
+	kvU32(&meta, "llama.context_length", 4096)
+	kvU32(&meta, "llama.attention.head_count", 32)
+	kvU32(&meta, "llama.feed_forward_length", 11008)
+	kvU32(&meta, "llama.block_count", 32)
+	kvU32(&meta, "tokenizer.ggml.bos_token_id", 1)
+	kvStr(&meta, "tokenizer.chat_template", chatTemplate)
+
+	// large array value — the one the optimization skips reading
+	wStr(&meta, "tokenizer.ggml.tokens")
+	binary.Write(&meta, binary.LittleEndian, ggufTypeArray)
+	binary.Write(&meta, binary.LittleEndian, ggufTypeString)
+	binary.Write(&meta, binary.LittleEndian, uint64(vocab))
+	for i := 0; i < vocab; i++ {
+		wStr(&meta, "token")
+	}
+
+	var out bytes.Buffer
+	binary.Write(&out, binary.LittleEndian, gguf.GGUFMagicGGUFLe)
+	binary.Write(&out, binary.LittleEndian, uint32(3)) // version
+	binary.Write(&out, binary.LittleEndian, uint64(0)) // tensor count
+	binary.Write(&out, binary.LittleEndian, uint64(9)) // metadata kv count
+	out.Write(meta.Bytes())
+
+	return os.WriteFile(path, out.Bytes(), 0o644)
+}
+
 var _ = Describe("Backend hooks and parser defaults", func() {
 	Context("MatchParserDefaults", func() {
 		It("matches Qwen3 family", func() {
@@ -137,6 +200,58 @@ var _ = Describe("Backend hooks and parser defaults", func() {
 		})
 	})
 
+	Context("llamaCppDefaults GGUF guessing", func() {
+		// Regression coverage for https://github.com/mudler/LocalAI/issues/9790:
+		// the hook reads GGUF headers with SkipLargeMetadata + UseMMap to avoid
+		// pulling the whole tokenizer vocab off (slow) disk on every startup. This
+		// verifies that skipping the vocab array still yields the correct guessed
+		// defaults from the remaining scalar metadata.
+		const chatTemplate = "{{ bos_token }}{% for m in messages %}{{ m.content }}{% endfor %}"
+
+		It("guesses defaults from a GGUF whose large vocab is skipped", func() {
+			dir := GinkgoT().TempDir()
+			modelFile := "repro.gguf"
+			Expect(writeTestGGUF(filepath.Join(dir, modelFile), chatTemplate, 50000)).To(Succeed())
+
+			// A pre-set context size short-circuits the GGUF run-estimate, which
+			// needs full tensor info this header-only fixture deliberately omits;
+			// the metadata-reading path the optimization touches is unaffected.
+			ctxSize := 4096
+			cfg := &ModelConfig{
+				Backend: "llama-cpp",
+				LLMConfig: LLMConfig{ContextSize: &ctxSize},
+				PredictionOptions: schema.PredictionOptions{
+					BasicModelRequest: schema.BasicModelRequest{Model: modelFile},
+				},
+			}
+			cfg.SetDefaults(ModelPath(dir))
+
+			// chat_template is a scalar string, not part of the skipped array,
+			// so it must be captured verbatim.
+			Expect(cfg.GetModelTemplate()).To(Equal(chatTemplate))
+			// scalar-derived defaults are still applied
+			Expect(cfg.ContextSize).NotTo(BeNil())
+			Expect(cfg.NGPULayers).NotTo(BeNil())
+			Expect(cfg.TemplateConfig.UseTokenizerTemplate).To(BeTrue())
+			Expect(cfg.KnownUsecaseStrings).To(ContainElement("FLAG_CHAT"))
+		})
+
+		It("falls back to the default context size when the GGUF is unreadable", func() {
+			dir := GinkgoT().TempDir()
+			Expect(os.WriteFile(filepath.Join(dir, "bad.gguf"), []byte("not a gguf"), 0o644)).To(Succeed())
+
+			cfg := &ModelConfig{
+				Backend: "llama-cpp",
+				PredictionOptions: schema.PredictionOptions{
+					BasicModelRequest: schema.BasicModelRequest{Model: "bad.gguf"},
+				},
+			}
+			cfg.SetDefaults(ModelPath(dir))
+
+			Expect(cfg.ContextSize).NotTo(BeNil())
+		})
+	})
+
 	Context("PromptCacheAll default", func() {
 		It("defaults to true when omitted from YAML", func() {
 			cfg := &ModelConfig{}

core/config/meta/registry.go

@@ -308,6 +308,41 @@ func DefaultRegistry() map[string]FieldMetaOverride {
 			},
 			Order: 64,
 		},
+		"pipeline.disable_thinking": {
+			Section:     "pipeline",
+			Label:       "Disable Thinking",
+			Description: "Suppress reasoning/thinking output from the pipeline LLM (sets enable_thinking=false on the underlying model). Use for models that emit <think> blocks you don't want spoken or streamed back to the realtime client.",
+			Component:   "toggle",
+			Order:       65,
+		},
+		"pipeline.streaming.llm": {
+			Section:     "pipeline",
+			Label:       "Stream LLM",
+			Description: "Stream LLM tokens to the realtime client as they are generated instead of waiting for the full response. Emits incremental response.output_audio_transcript.delta / text deltas.",
+			Component:   "toggle",
+			Order:       66,
+		},
+		"pipeline.streaming.tts": {
+			Section:     "pipeline",
+			Label:       "Stream TTS",
+			Description: "Stream synthesized audio chunks to the realtime client as they are produced (requires a TTS backend that implements TTSStream). Falls back to unary synthesis otherwise.",
+			Component:   "toggle",
+			Order:       67,
+		},
+		"pipeline.streaming.transcription": {
+			Section:     "pipeline",
+			Label:       "Stream Transcription",
+			Description: "Stream partial transcription text to the realtime client as the STT backend produces it (requires a transcription backend that implements AudioTranscriptionStream). Falls back to unary transcription otherwise.",
+			Component:   "toggle",
+			Order:       68,
+		},
+		"pipeline.streaming.clause_chunking": {
+			Section:     "pipeline",
+			Label:       "Clause Chunking",
+			Description: "Split the streamed reply into speakable clauses and synthesize each as soon as it completes, instead of buffering the whole message before TTS — lower time-to-first-audio. Script-aware (handles CJK 。！？ and Thai/Lao spaces), so it does not whitespace-split. Requires Stream LLM; off buffers the whole message.",
+			Component:   "toggle",
+			Order:       69,
+		},
 
 		// --- Functions ---
 		"function.grammar.parallel_calls": {

core/config/model_config.go

@@ -499,6 +499,16 @@ type Pipeline struct {
 	// the pipeline's LLM without editing the LLM model config. Overrides the LLM's
 	// own reasoning_effort. Unset leaves the LLM model config in charge.
 	ReasoningEffort string `yaml:"reasoning_effort,omitempty" json:"reasoning_effort,omitempty"`
+
+	// Streaming opts each pipeline stage into incremental delivery (LLM tokens,
+	// TTS audio chunks, transcription text). Unset stages keep the blocking
+	// unary path, so existing configs are unaffected.
+	Streaming PipelineStreaming `yaml:"streaming,omitempty" json:"streaming,omitempty"`
+
+	// DisableThinking suppresses reasoning/thinking for the pipeline LLM (maps
+	// to enable_thinking=false backend metadata) without editing the underlying
+	// LLM model config. Unset leaves the LLM model config in charge.
+	DisableThinking *bool `yaml:"disable_thinking,omitempty" json:"disable_thinking,omitempty"`
 }
 
 // ApplyReasoningEffort resolves the effective reasoning effort — a per-request
@@ -530,6 +540,41 @@ func (c *ModelConfig) ApplyReasoningEffort(requestEffort string) {
 	}
 }
 
+// @Description PipelineStreaming toggles incremental delivery per realtime stage.
+type PipelineStreaming struct {
+	LLM           *bool `yaml:"llm,omitempty" json:"llm,omitempty"`
+	TTS           *bool `yaml:"tts,omitempty" json:"tts,omitempty"`
+	Transcription *bool `yaml:"transcription,omitempty" json:"transcription,omitempty"`
+	// ClauseChunking splits the streamed LLM reply into speakable clauses and
+	// synthesizes each as soon as it completes, instead of buffering the whole
+	// message before TTS. Script-aware (CJK/Thai), so it does not rely on
+	// whitespace sentence boundaries. Requires LLM streaming; unset buffers the
+	// whole message (today's default).
+	ClauseChunking *bool `yaml:"clause_chunking,omitempty" json:"clause_chunking,omitempty"`
+}
+
+// StreamLLM reports whether LLM tokens should be streamed for this pipeline.
+func (p Pipeline) StreamLLM() bool { return p.Streaming.LLM != nil && *p.Streaming.LLM }
+
+// StreamTTS reports whether TTS audio should be streamed for this pipeline.
+func (p Pipeline) StreamTTS() bool { return p.Streaming.TTS != nil && *p.Streaming.TTS }
+
+// StreamTranscription reports whether transcription text should be streamed.
+func (p Pipeline) StreamTranscription() bool {
+	return p.Streaming.Transcription != nil && *p.Streaming.Transcription
+}
+
+// ChunkClauses reports whether the streamed reply should be split into
+// script-aware clauses and synthesized incrementally rather than buffered whole.
+func (p Pipeline) ChunkClauses() bool {
+	return p.Streaming.ClauseChunking != nil && *p.Streaming.ClauseChunking
+}
+
+// ThinkingDisabled reports whether the pipeline forces the LLM's thinking off.
+func (p Pipeline) ThinkingDisabled() bool {
+	return p.DisableThinking != nil && *p.DisableThinking
+}
+
 // @Description File configuration for model downloads
 type File struct {
 	Filename string         `yaml:"filename,omitempty" json:"filename,omitempty"`

core/config/mtp.go

@@ -30,11 +30,26 @@ func MTPSpecOptions() []string {
 	return out
 }
 
-// HasEmbeddedMTPHead reports whether the parsed GGUF declares a Multi-Token
-// Prediction head. Detection reads `<arch>.nextn_predict_layers`, which is
-// what `gguf_writer.add_nextn_predict_layers(n)` emits in upstream's
+// isDraftOnlyAssistantArch reports whether an architecture names a standalone
+// MTP *draft* model rather than a self-speculating trunk. Upstream's Gemma4 MTP
+// (ggml-org/llama.cpp#23398) registers the head as a separate `gemma4-assistant`
+// architecture whose GGUF still carries `nextn_predict_layers`, but which cannot
+// run alone: it requires a paired target context (`ctx_other`). Such archs must
+// not trigger the embedded-head self-speculation defaults. The `-assistant`
+// suffix is upstream's naming convention for these draft-only checkpoints.
+func isDraftOnlyAssistantArch(arch string) bool {
+	return strings.HasSuffix(arch, "-assistant")
+}
+
+// HasEmbeddedMTPHead reports whether the parsed GGUF declares a self-speculating
+// Multi-Token Prediction head. Detection reads `<arch>.nextn_predict_layers`,
+// which is what `gguf_writer.add_nextn_predict_layers(n)` emits in upstream's
 // `conversion/qwen.py` MTP mixin. A positive layer count means the head is
 // present in the same GGUF as the trunk.
+//
+// Draft-only assistant architectures (e.g. Gemma4's `gemma4-assistant`) carry
+// the same key but are separate draft checkpoints meant to be paired with a
+// target model, so they are deliberately excluded here.
 func HasEmbeddedMTPHead(f *gguf.GGUFFile) (uint32, bool) {
 	if f == nil {
 		return 0, false
@@ -43,6 +58,9 @@ func HasEmbeddedMTPHead(f *gguf.GGUFFile) (uint32, bool) {
 	if arch == "" {
 		return 0, false
 	}
+	if isDraftOnlyAssistantArch(arch) {
+		return 0, false
+	}
 	v, ok := f.Header.MetadataKV.Get(arch + ".nextn_predict_layers")
 	if !ok {
 		return 0, false

core/config/mtp_test.go

@@ -3,10 +3,33 @@ package config_test
 import (
 	. "github.com/mudler/LocalAI/core/config"
 
+	gguf "github.com/gpustack/gguf-parser-go"
 	. "github.com/onsi/ginkgo/v2"
 	. "github.com/onsi/gomega"
 )
 
+// ggufWithArch fabricates a minimal in-memory GGUF carrying the given
+// `general.architecture` and a positive `<arch>.nextn_predict_layers` count,
+// so HasEmbeddedMTPHead can be exercised without a real model file.
+func ggufWithArch(arch string, nextn uint32) *gguf.GGUFFile {
+	return &gguf.GGUFFile{
+		Header: gguf.GGUFHeader{
+			MetadataKV: gguf.GGUFMetadataKVs{
+				{
+					Key:       "general.architecture",
+					ValueType: gguf.GGUFMetadataValueTypeString,
+					Value:     arch,
+				},
+				{
+					Key:       arch + ".nextn_predict_layers",
+					ValueType: gguf.GGUFMetadataValueTypeUint32,
+					Value:     nextn,
+				},
+			},
+		},
+	}
+}
+
 var _ = Describe("MTP auto-defaults", func() {
 	Context("MTPSpecOptions", func() {
 		It("returns the upstream-recommended speculative tuple", func() {
@@ -82,5 +105,20 @@ var _ = Describe("MTP auto-defaults", func() {
 			Expect(ok).To(BeFalse())
 			Expect(n).To(BeZero())
 		})
+
+		It("detects a same-GGUF embedded head (DeepSeek/Qwen style)", func() {
+			n, ok := HasEmbeddedMTPHead(ggufWithArch("qwen3moe", 1))
+			Expect(ok).To(BeTrue())
+			Expect(n).To(Equal(uint32(1)))
+		})
+
+		It("ignores a gemma4-assistant draft-only model", func() {
+			// The assistant GGUF carries nextn_predict_layers but is a separate
+			// draft model that requires a paired target (ctx_other); it cannot
+			// self-speculate, so it must not trigger the embedded-head defaults.
+			n, ok := HasEmbeddedMTPHead(ggufWithArch("gemma4-assistant", 48))
+			Expect(ok).To(BeFalse())
+			Expect(n).To(BeZero())
+		})
 	})
 })

core/config/pipeline_streaming_test.go

@@ -0,0 +1,57 @@
+package config
+
+import (
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+	"gopkg.in/yaml.v3"
+)
+
+// The realtime pipeline can stream each stage (LLM tokens, TTS audio,
+// transcription text) and can disable model "thinking" for the LLM. These are
+// opt-in per pipeline; everything defaults to off so existing configs keep the
+// unary behaviour.
+var _ = Describe("Pipeline streaming config", func() {
+	It("defaults every streaming + thinking helper to false when unset", func() {
+		var p Pipeline
+		Expect(p.StreamLLM()).To(BeFalse())
+		Expect(p.StreamTTS()).To(BeFalse())
+		Expect(p.StreamTranscription()).To(BeFalse())
+		Expect(p.ChunkClauses()).To(BeFalse())
+		Expect(p.ThinkingDisabled()).To(BeFalse())
+	})
+
+	It("parses the nested streaming block and disable_thinking from YAML", func() {
+		var c ModelConfig
+		err := yaml.Unmarshal([]byte(`
+name: gpt-realtime
+pipeline:
+  llm: my-llm
+  tts: my-tts
+  transcription: my-stt
+  streaming:
+    llm: true
+    tts: true
+    transcription: true
+    clause_chunking: true
+  disable_thinking: true
+`), &c)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(c.Pipeline.StreamLLM()).To(BeTrue())
+		Expect(c.Pipeline.StreamTTS()).To(BeTrue())
+		Expect(c.Pipeline.StreamTranscription()).To(BeTrue())
+		Expect(c.Pipeline.ChunkClauses()).To(BeTrue())
+		Expect(c.Pipeline.ThinkingDisabled()).To(BeTrue())
+	})
+
+	It("treats an explicit false in the streaming block as disabled", func() {
+		var c ModelConfig
+		err := yaml.Unmarshal([]byte(`
+name: gpt-realtime
+pipeline:
+  streaming:
+    tts: false
+`), &c)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(c.Pipeline.StreamTTS()).To(BeFalse())
+	})
+})

core/http/endpoints/openai/chat.go

@@ -103,7 +103,12 @@ func applyAutoparserOverride(
 	// blocks like "<think></think>" that some models emit when reasoning
 	// is disabled.
 	if deltaReasoning == "" && deltaContent != "" {
-		deltaReasoning, deltaContent = reason.ExtractReasoningWithConfig(deltaContent, thinkingStartToken, reasoningConfig)
+		// Complete-response extraction: only honor a prefilled <think> start
+		// token when deltaContent actually closes the reasoning block. Without
+		// it the model answered directly and the whole answer must stay in
+		// content rather than be swallowed as unclosed reasoning. See
+		// reason.ExtractReasoningComplete.
+		deltaReasoning, deltaContent = reason.ExtractReasoningComplete(deltaContent, thinkingStartToken, reasoningConfig)
 	}
 	xlog.Debug("[ChatDeltas] non-SSE no-tools: overriding result with C++ autoparser deltas",
 		"content_len", len(deltaContent), "reasoning_len", len(deltaReasoning))

core/http/endpoints/openai/chat_test.go

@@ -186,6 +186,114 @@ var _ = Describe("applyAutoparserOverride", func() {
 			Expect(result).To(Equal(existing))
 		})
 	})
+
+	// Regression tests for the prefilled-thinking-token path (thinkingStartToken
+	// != ""). This is the configuration the gallery qwen3 family runs in: the
+	// chat template injects <think> into the prompt, so DetectThinkingStartToken
+	// returns "<think>" and the model's output begins *inside* a reasoning block
+	// — it emits a closing </think> but no opening tag.
+	//
+	// The defensive Go-side fallback prepends the start token so the standard
+	// extractor can pair it with the model's </think>. But on a *complete*
+	// response that contains NO closing tag (the model answered directly with no
+	// reasoning at all), prepending <think> manufactures an unclosed block that
+	// swallows the entire answer into reasoning, leaving content empty. That is
+	// the bug: short/direct answers (session names, JSON summaries) come back
+	// with an empty content field.
+	Context("autoparser delivered content with empty reasoning and a prefilled thinking token", func() {
+		const startToken = "<think>"
+
+		It("keeps a tag-less direct answer as content instead of swallowing it as reasoning", func() {
+			// Model answered directly: no <think>, no </think> anywhere.
+			chatDeltas := []*pb.ChatDelta{
+				{Content: "hello", ReasoningContent: ""},
+			}
+
+			result := applyAutoparserOverride(chatDeltas, startToken, reason.Config{}, nil)
+
+			Expect(result).To(HaveLen(1))
+			Expect(result[0].Message.Content).ToNot(BeNil())
+			Expect(*(result[0].Message.Content.(*string))).To(Equal("hello"),
+				"a complete answer with no closing reasoning tag must stay in content")
+			Expect(result[0].Message.Reasoning).To(BeNil(),
+				"no reasoning block was emitted, so Reasoning must not be set")
+		})
+
+		It("keeps a tag-less JSON answer as content (the summary case)", func() {
+			raw := `{"short":"Tests pass","long":"go test ./... succeeded."}`
+			chatDeltas := []*pb.ChatDelta{
+				{Content: raw, ReasoningContent: ""},
+			}
+
+			result := applyAutoparserOverride(chatDeltas, startToken, reason.Config{}, nil)
+
+			Expect(result).To(HaveLen(1))
+			Expect(*(result[0].Message.Content.(*string))).To(Equal(raw))
+			Expect(result[0].Message.Reasoning).To(BeNil())
+		})
+
+		It("still splits reasoning when the model emits the closing tag (prefill paired with </think>)", func() {
+			// The legitimate prefill case: <think> was in the prompt, so the
+			// output carries only the closing tag. The closing tag is the proof
+			// that a reasoning block exists, so extraction must run.
+			raw := "The user wants a greeting.\n</think>\n\nHello there!"
+			chatDeltas := []*pb.ChatDelta{
+				{Content: raw, ReasoningContent: ""},
+			}
+
+			result := applyAutoparserOverride(chatDeltas, startToken, reason.Config{}, nil)
+
+			Expect(result).To(HaveLen(1))
+			content := *(result[0].Message.Content.(*string))
+			Expect(content).To(ContainSubstring("Hello there!"))
+			Expect(content).ToNot(ContainSubstring("</think>"))
+			Expect(content).ToNot(ContainSubstring("The user wants a greeting"))
+			Expect(result[0].Message.Reasoning).ToNot(BeNil())
+			Expect(*result[0].Message.Reasoning).To(ContainSubstring("The user wants a greeting"))
+		})
+
+		It("still splits a fully-tagged <think>…</think> block with a prefill token set", func() {
+			raw := "<think>Reasoning here.</think>Final answer."
+			chatDeltas := []*pb.ChatDelta{
+				{Content: raw, ReasoningContent: ""},
+			}
+
+			result := applyAutoparserOverride(chatDeltas, startToken, reason.Config{}, nil)
+
+			Expect(result).To(HaveLen(1))
+			Expect(*(result[0].Message.Content.(*string))).To(Equal("Final answer."))
+			Expect(result[0].Message.Reasoning).ToNot(BeNil())
+			Expect(*result[0].Message.Reasoning).To(ContainSubstring("Reasoning here"))
+		})
+
+		// End-to-end regression for the real production failure: a request with
+		// enable_thinking=false against a <think>-capable model (qwen3 family).
+		//
+		// In non-thinking mode the model emits no reasoning block, so llama.cpp's
+		// autoparser correctly returns ChatDeltas with Content set and
+		// ReasoningContent EMPTY (verified against stock llama-server: the same
+		// model with chat_template_kwargs.enable_thinking=false returns
+		// reasoning_content=null and content="hello"). But thinkingStartToken is
+		// detected per-model from the enable_thinking=TRUE render
+		// (grpc-server renders with enable_thinking=true; DetectThinkingStartToken
+		// does not evaluate the jinja {% if enable_thinking %} conditional), so it
+		// is "<think>" even for this non-thinking request. The old code prepended
+		// it and swallowed the answer. This is the case that broke session
+		// summaries and auto-titles and was NOT covered before.
+		It("preserves content for a non-thinking-mode request (enable_thinking=false, empty reasoning_content)", func() {
+			// What llama.cpp's autoparser actually returns in non-thinking mode.
+			chatDeltas := []*pb.ChatDelta{
+				{Content: `{"short":"Go tests passed for internal/session"}`, ReasoningContent: ""},
+			}
+
+			result := applyAutoparserOverride(chatDeltas, startToken, reason.Config{}, nil)
+
+			Expect(result).To(HaveLen(1))
+			Expect(*(result[0].Message.Content.(*string))).To(Equal(`{"short":"Go tests passed for internal/session"}`),
+				"non-thinking-mode answers must reach the client intact, not be swallowed as reasoning")
+			Expect(result[0].Message.Reasoning).To(BeNil())
+		})
+	})
 })
 
 var _ = Describe("mergeToolCallDeltas", func() {

core/http/endpoints/openai/realtime.go

@@ -2,8 +2,10 @@ package openai
 
 import (
 	"context"
+	"crypto/rand"
 	"encoding/base64"
 	"encoding/binary"
+	"encoding/hex"
 	"encoding/json"
 	"fmt"
 	"math"
@@ -235,6 +237,12 @@ type Model interface {
 	Transcribe(ctx context.Context, audio, language string, translate bool, diarize bool, prompt string) (*schema.TranscriptionResult, error)
 	Predict(ctx context.Context, messages schema.Messages, images, videos, audios []string, tokenCallback func(string, backend.TokenUsage) bool, tools []types.ToolUnion, toolChoice *types.ToolChoiceUnion, logprobs *int, topLogprobs *int, logitBias map[string]float64) (func() (backend.LLMResponse, error), error)
 	TTS(ctx context.Context, text, voice, language string) (string, *proto.Result, error)
+	// TTSStream synthesizes speech incrementally, invoking onAudio with raw PCM
+	// chunks (and the backend sample rate) as they are produced.
+	TTSStream(ctx context.Context, text, voice, language string, onAudio func(pcm []byte, sampleRate int) error) error
+	// TranscribeStream transcribes audio incrementally, invoking onDelta for each
+	// transcript text fragment and returning the final aggregated result.
+	TranscribeStream(ctx context.Context, audio, language string, translate, diarize bool, prompt string, onDelta func(text string)) (*schema.TranscriptionResult, error)
 	PredictConfig() *config.ModelConfig
 }
 
@@ -1254,27 +1262,15 @@ func commitUtterance(ctx context.Context, utt []byte, session *Session, conv *Co
 	// TODO: If we have a real any-to-any model then transcription is optional
 	var transcript string
 	if session.InputAudioTranscription != nil {
-		tr, err := session.ModelInterface.Transcribe(ctx, f.Name(), session.InputAudioTranscription.Language, false, false, session.InputAudioTranscription.Prompt)
+		// emitTranscription streams transcript deltas when
+		// pipeline.streaming.transcription is set, otherwise emits a single
+		// completed event; either way it returns the final transcript text.
+		var err error
+		transcript, err = emitTranscription(ctx, t, session, generateItemID(), f.Name())
 		if err != nil {
 			sendError(t, "transcription_failed", err.Error(), "", "event_TODO")
 			return
-		} else if tr == nil {
-			sendError(t, "transcription_failed", "trancribe result is nil", "", "event_TODO")
-			return
 		}
-
-		transcript = tr.Text
-		sendEvent(t, types.ConversationItemInputAudioTranscriptionCompletedEvent{
-			ServerEventBase: types.ServerEventBase{
-				EventID: "event_TODO",
-			},
-
-			ItemID: generateItemID(),
-			// ResponseID:   "resp_TODO", // Not needed for transcription completed event
-			// OutputIndex:  0,
-			ContentIndex: 0,
-			Transcript:   transcript,
-		})
 	} else {
 		sendNotImplemented(t, "any-to-any models")
 		return
@@ -1502,6 +1498,26 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 		},
 	})
 
+	// Streamed LLM path: when the pipeline opts into LLM streaming, stream the
+	// transcript to the client as it is generated and synthesize the buffered
+	// message once. Tool turns are supported only when the model uses its
+	// tokenizer template: the C++ autoparser then delivers content and tool
+	// calls via ChatDeltas (clearing the text stream), so the spoken transcript
+	// never leaks tool-call tokens. Grammar-based function calling emits the
+	// call as JSON in the token stream, so those turns keep the buffered path.
+	if config != nil && session.ModelConfig != nil && session.ModelConfig.Pipeline.StreamLLM() {
+		canStream := len(tools) == 0 || config.TemplateConfig.UseTokenizerTemplate
+		var respMods []types.Modality
+		if overrides != nil {
+			respMods = overrides.OutputModalities
+		}
+		if canStream && modalitiesContainAudio(resolveOutputModalities(session.OutputModalities, respMods)) {
+			if streamLLMResponse(ctx, session, conv, t, responseID, conversationHistory, images, config, tools, toolChoice, toolTurn) {
+				return
+			}
+		}
+	}
+
 	predFunc, err := session.ModelInterface.Predict(ctx, conversationHistory, images, nil, nil, nil, tools, toolChoice, nil, nil, nil)
 	if err != nil {
 		sendError(t, "inference_failed", fmt.Sprintf("backend error: %v", err), "", "") // item.Assistant.ID is unknown here
@@ -1579,7 +1595,7 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 		// ExtractReasoningWithConfig is a no-op when no tag pair matches,
 		// so it's safe to apply unconditionally in the no-reasoning branch.
 		if deltaReasoning == "" && deltaContent != "" {
-			deltaReasoning, deltaContent = reasoning.ExtractReasoningWithConfig(deltaContent, thinkingStartToken, config.ReasoningConfig)
+			deltaReasoning, deltaContent = reasoning.ExtractReasoningComplete(deltaContent, thinkingStartToken, spokenReasoningConfig(config.ReasoningConfig))
 		}
 		reasoningText = deltaReasoning
 		responseWithoutReasoning = deltaContent
@@ -1587,7 +1603,7 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 		cleanedResponse = deltaContent
 		toolCalls = deltaToolCalls
 	} else {
-		reasoningText, responseWithoutReasoning = reasoning.ExtractReasoningWithConfig(rawResponse, thinkingStartToken, config.ReasoningConfig)
+		reasoningText, responseWithoutReasoning = reasoning.ExtractReasoningComplete(rawResponse, thinkingStartToken, spokenReasoningConfig(config.ReasoningConfig))
 		textContent = functions.ParseTextContent(responseWithoutReasoning, config.FunctionsConfig)
 		cleanedResponse = functions.CleanupLLMResult(responseWithoutReasoning, config.FunctionsConfig)
 		toolCalls = functions.ParseFunctionCall(cleanedResponse, config.FunctionsConfig)
@@ -1713,64 +1729,7 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 				return
 			}
 
-			audioFilePath, res, err := session.ModelInterface.TTS(ctx, finalSpeech, session.Voice, session.InputAudioTranscription.Language)
-			if err != nil {
-				if ctx.Err() != nil {
-					xlog.Debug("TTS cancelled (barge-in)")
-					sendCancelledResponse()
-					return
-				}
-				xlog.Error("TTS failed", "error", err)
-				sendError(t, "tts_error", fmt.Sprintf("TTS generation failed: %v", err), "", item.Assistant.ID)
-				return
-			}
-			if !res.Success {
-				xlog.Error("TTS failed", "message", res.Message)
-				sendError(t, "tts_error", fmt.Sprintf("TTS generation failed: %s", res.Message), "", item.Assistant.ID)
-				return
-			}
-			defer func() { _ = os.Remove(audioFilePath) }()
-
-			audioBytes, err := os.ReadFile(audioFilePath)
-			if err != nil {
-				xlog.Error("failed to read TTS file", "error", err)
-				sendError(t, "tts_error", fmt.Sprintf("Failed to read TTS audio: %v", err), "", item.Assistant.ID)
-				return
-			}
-
-			// Parse WAV header to get raw PCM and the actual sample rate from the TTS backend.
-			pcmData, ttsSampleRate := laudio.ParseWAV(audioBytes)
-			if ttsSampleRate == 0 {
-				ttsSampleRate = localSampleRate
-			}
-			xlog.Debug("TTS audio parsed", "raw_bytes", len(audioBytes), "pcm_bytes", len(pcmData), "sample_rate", ttsSampleRate)
-
-			// SendAudio (WebRTC) passes PCM at the TTS sample rate directly to the
-			// Opus encoder, which resamples to 48kHz internally. This avoids a
-			// lossy intermediate resample through 16kHz.
-			// XXX: This is a noop in websocket mode; it's included in the JSON instead
-			if err := t.SendAudio(ctx, pcmData, ttsSampleRate); err != nil {
-				if ctx.Err() != nil {
-					xlog.Debug("Audio playback cancelled (barge-in)")
-					sendCancelledResponse()
-					return
-				}
-				xlog.Error("failed to send audio via transport", "error", err)
-			}
-
-			// For WebSocket clients, resample to the session's output rate and
-			// deliver audio as base64 in JSON events. WebRTC clients already
-			// received audio over the RTP track, so skip the base64 payload.
-			if !isWebRTC {
-				wsPCM := pcmData
-				if ttsSampleRate != session.OutputSampleRate {
-					samples := sound.BytesToInt16sLE(pcmData)
-					resampled := sound.ResampleInt16(samples, ttsSampleRate, session.OutputSampleRate)
-					wsPCM = sound.Int16toBytesLE(resampled)
-				}
-				audioString = base64.StdEncoding.EncodeToString(wsPCM)
-			}
-
+			// Transcript of the spoken reply (the audio's text).
 			sendEvent(t, types.ResponseOutputAudioTranscriptDeltaEvent{
 				ServerEventBase: types.ServerEventBase{},
 				ResponseID:      responseID,
@@ -1788,15 +1747,26 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 				Transcript:      finalSpeech,
 			})
 
+			// Synthesize and send the audio. With pipeline.streaming.tts enabled
+			// emitSpeech forwards a response.output_audio.delta per backend PCM
+			// chunk as it's produced; otherwise it sends the whole utterance as a
+			// single delta. The returned PCM is stored (base64) on the item below.
+			pcmAudio, err := emitSpeech(ctx, t, session, responseID, item.Assistant.ID, finalSpeech)
+			if err != nil {
+				if ctx.Err() != nil {
+					xlog.Debug("TTS cancelled (barge-in)")
+					sendCancelledResponse()
+					return
+				}
+				xlog.Error("TTS failed", "error", err)
+				sendError(t, "tts_error", fmt.Sprintf("TTS generation failed: %v", err), "", item.Assistant.ID)
+				return
+			}
+			if !isWebRTC {
+				audioString = base64.StdEncoding.EncodeToString(pcmAudio)
+			}
+
 			if !isWebRTC {
-				sendEvent(t, types.ResponseOutputAudioDeltaEvent{
-					ServerEventBase: types.ServerEventBase{},
-					ResponseID:      responseID,
-					ItemID:          item.Assistant.ID,
-					OutputIndex:     0,
-					ContentIndex:    0,
-					Delta:           audioString,
-				})
 				sendEvent(t, types.ResponseOutputAudioDoneEvent{
 					ServerEventBase: types.ServerEventBase{},
 					ResponseID:      responseID,
@@ -1849,17 +1819,27 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 		})
 	}
 
-	// Handle Tool Calls. Two paths:
-	//   - LocalAI Assistant tools (session.AssistantExecutor.IsTool) run
-	//     server-side; we append both the call and its output to conv.Items
-	//     and re-trigger a follow-up response so the model can speak the
-	//     result. The client only sees observability events.
-	//   - All other tools follow the standard OpenAI flow: emit
-	//     function_call_arguments.done and wait for the client to send
-	//     conversation.item.create back.
-	xlog.Debug("About to handle tool calls", "finalToolCallsCount", len(finalToolCalls))
+	// Emit the parsed tool calls, the terminal response.done, and (for
+	// server-side assistant tools) the follow-up response. Shared with the
+	// streamed path so both finalize tool calls identically.
+	emitToolCallItems(ctx, session, conv, t, responseID, finalToolCalls, finalSpeech != "", toolTurn)
+}
+
+// emitToolCallItems emits the realtime function_call items for the parsed tool
+// calls, the terminal response.done, and — for server-side LocalAI Assistant
+// tools — re-triggers a follow-up response so the model can speak the result.
+// hasContent shifts the tool-call output index past the assistant content item
+// when the same turn also produced spoken/text content. Two tool paths:
+//   - LocalAI Assistant tools (session.AssistantExecutor.IsTool) run server-side;
+//     we append both the call and its output to conv.Items and re-trigger. The
+//     client only sees observability events.
+//   - All other tools follow the standard OpenAI flow: emit
+//     function_call_arguments.done and wait for the client to send
+//     conversation.item.create back.
+func emitToolCallItems(ctx context.Context, session *Session, conv *Conversation, t Transport, responseID string, toolCalls []functions.FuncCallResults, hasContent bool, toolTurn int) {
+	xlog.Debug("About to handle tool calls", "finalToolCallsCount", len(toolCalls))
 	executedAssistantTool := false
-	for i, tc := range finalToolCalls {
+	for i, tc := range toolCalls {
 		toolCallID := generateItemID()
 		callID := "call_" + generateUniqueID() // OpenAI uses call_xyz
 
@@ -1879,7 +1859,7 @@ func triggerResponseAtTurn(ctx context.Context, session *Session, conv *Conversa
 		conv.Lock.Unlock()
 
 		outputIndex := i
-		if finalSpeech != "" {
+		if hasContent {
 			outputIndex++
 		}
 
@@ -2005,8 +1985,11 @@ func generateItemID() string {
 }
 
 func generateUniqueID() string {
-	// Generate a unique ID string
-	// For simplicity, use a counter or UUID
-	// Implement as needed
-	return "unique_id"
+	// 16 random bytes, hex-encoded. Must be collision-free: session, item,
+	// response and call IDs build on this, and the conversation tracks/removes
+	// items by ID (e.g. cancel() in realtime_stream.go, conversation.item.retrieve).
+	// A constant would make every ID alias and corrupt that bookkeeping.
+	var b [16]byte
+	_, _ = rand.Read(b[:])
+	return hex.EncodeToString(b[:])
 }

core/http/endpoints/openai/realtime_chunker.go

@@ -0,0 +1,200 @@
+package openai
+
+import (
+	"strings"
+	"unicode"
+	"unicode/utf8"
+
+	"github.com/rivo/uniseg"
+)
+
+// Default clause-chunker bounds (in runes). minRunes gates only sub-sentence
+// (clause-mark / Thai-space) cuts so we don't synthesize tiny choppy fragments;
+// full sentences always flush regardless of length. maxRunes caps an
+// unterminated run so a long punctuation-less span doesn't buffer unbounded.
+const (
+	defaultClauseMinRunes = 12
+	defaultClauseMaxRunes = 200
+)
+
+// clauseChunker splits streamed LLM content into speakable clauses for
+// incremental TTS, in a SCRIPT-AWARE way so it works for languages without
+// whitespace word boundaries. It leans on UAX #29 sentence segmentation (which
+// natively terminates on CJK 。！？ as well as Latin .!?), adds CJK clause
+// punctuation (，、；：) and Thai/Lao spaces as finer boundaries, and caps an
+// over-long unterminated run via UAX #14 line-break opportunities.
+//
+// Unlike the old ASCII .!?/newline segmenter (dropped in 076dcdbe), it does not
+// degrade to whole-message buffering for CJK (handled natively) or Thai/Lao
+// (handled via spaces, which Thai uses at clause/sentence boundaries). Scripts
+// that genuinely need a dictionary (Khmer/Myanmar) simply stay buffered until a
+// space or end-of-message — no worse than the buffered default.
+//
+// It is not safe for concurrent use; callers feed it from a single goroutine
+// (the LLM token callback).
+type clauseChunker struct {
+	buf      strings.Builder
+	minRunes int
+	maxRunes int
+}
+
+func newClauseChunker(minRunes, maxRunes int) *clauseChunker {
+	return &clauseChunker{minRunes: minRunes, maxRunes: maxRunes}
+}
+
+// push appends streamed content and returns any clauses that are now complete —
+// "complete" meaning confirmed by following content, so we never speak a clause
+// that the next token might extend. Incomplete trailing text stays buffered.
+func (c *clauseChunker) push(text string) []string {
+	c.buf.WriteString(text)
+	return c.drain(false)
+}
+
+// flush returns the remaining buffered clauses, treating end-of-input as a hard
+// boundary, and clears the buffer.
+func (c *clauseChunker) flush() []string {
+	return c.drain(true)
+}
+
+func (c *clauseChunker) drain(final bool) []string {
+	s := c.buf.String()
+	rest := s
+	var out []string
+	for rest != "" {
+		end, ok := c.nextBoundary(rest, final)
+		if !ok {
+			break
+		}
+		if seg := strings.TrimSpace(rest[:end]); seg != "" {
+			out = append(out, seg)
+		}
+		rest = rest[end:]
+	}
+	// Rewriting the builder reallocates and copies the whole buffer; skip it on
+	// the common per-token call where no boundary was confirmed.
+	if len(rest) != len(s) {
+		c.buf.Reset()
+		c.buf.WriteString(rest)
+	}
+	return out
+}
+
+// nextBoundary returns the byte offset just past the first emittable clause in
+// s, or ok=false when more input is needed (final=false) and no boundary is
+// confirmed yet.
+func (c *clauseChunker) nextBoundary(s string, final bool) (int, bool) {
+	if s == "" {
+		return 0, false
+	}
+
+	// 1) UAX #29 sentence boundary. When the first sentence is followed by more
+	//    text it is a confirmed complete sentence (handles Latin .!? with
+	//    abbreviation/decimal guards, and CJK 。！？ with no whitespace).
+	sentence, rest, _ := uniseg.FirstSentenceInString(s, -1)
+	if rest != "" {
+		// Optionally cut finer inside the sentence at a clause boundary.
+		if cut, ok := c.firstClauseCut(sentence); ok {
+			return cut, true
+		}
+		return len(sentence), true
+	}
+
+	// 2) Unterminated tail: look for a sub-sentence clause boundary (CJK
+	//    punctuation or a Thai/Lao space) confirmed by following content.
+	if cut, ok := c.firstClauseCut(s); ok {
+		return cut, true
+	}
+
+	// 3) Over-long punctuation-less run: force a typographically legal break so
+	//    we don't buffer unbounded (e.g. a long CJK run with no punctuation).
+	if !final && c.maxRunes > 0 && utf8.RuneCountInString(s) > c.maxRunes {
+		if cut, ok := lineBreakCut(s, c.maxRunes); ok {
+			return cut, true
+		}
+	}
+
+	// 4) End of input: emit whatever remains as the final clause.
+	if final {
+		return len(s), true
+	}
+	return 0, false
+}
+
+// firstClauseCut returns the byte offset just past the first sub-sentence clause
+// boundary in s — a CJK clause punctuation mark, or a space following a Thai/Lao
+// letter — provided the prefix is at least minRunes long and non-space content
+// follows. The boundary mark (and any trailing spaces) stay with the left clause.
+func (c *clauseChunker) firstClauseCut(s string) (int, bool) {
+	var prev rune
+	runes := 0
+	for i, r := range s {
+		boundary := isCJKClausePunct(r) || (unicode.IsSpace(r) && isThaiLao(prev))
+		if boundary && runes+1 >= c.minRunes {
+			end := i + utf8.RuneLen(r)
+			for end < len(s) {
+				nr, sz := utf8.DecodeRuneInString(s[end:])
+				if !unicode.IsSpace(nr) {
+					break
+				}
+				end += sz
+			}
+			if end < len(s) { // confirmed: real content follows the boundary
+				return end, true
+			}
+			// Boundary sits at the end of the buffer with nothing after it yet —
+			// wait for the next token to confirm it rather than emit early.
+			return 0, false
+		}
+		prev = r
+		runes++
+	}
+	return 0, false
+}
+
+// lineBreakCut walks UAX #14 line segments and returns the byte offset of the
+// last legal break opportunity at or before maxRunes. Returns ok=false when the
+// run has no internal break opportunity (e.g. a space-less Thai run), leaving it
+// buffered.
+func lineBreakCut(s string, maxRunes int) (int, bool) {
+	state := -1
+	rest := s
+	consumed := 0
+	runes := 0
+	for rest != "" {
+		seg, rem, _, st := uniseg.FirstLineSegmentInString(rest, state)
+		state = st
+		runes += utf8.RuneCountInString(seg)
+		consumed += len(seg)
+		rest = rem
+		if runes >= maxRunes {
+			if consumed < len(s) {
+				return consumed, true
+			}
+			return 0, false
+		}
+	}
+	return 0, false
+}
+
+// isCJKClausePunct reports whether r is a CJK clause-level separator worth a
+// soft TTS break. Sentence terminators (。！？) are intentionally excluded — UAX
+// #29 sentence segmentation already handles those.
+func isCJKClausePunct(r rune) bool {
+	switch r {
+	case '，', // ， fullwidth comma
+		'、', // 、 ideographic comma
+		'；', // ； fullwidth semicolon
+		'：', // ： fullwidth colon
+		'・', // ・ katakana middle dot
+		'･': // ・ halfwidth katakana middle dot
+		return true
+	}
+	return false
+}
+
+// isThaiLao reports whether r is a Thai or Lao letter. Those scripts have no
+// inter-word spaces; an ASCII space inside such a run marks a clause/sentence
+// boundary, which is the only no-dictionary segmentation signal available.
+func isThaiLao(r rune) bool {
+	return unicode.Is(unicode.Thai, r) || unicode.Is(unicode.Lao, r)
+}

core/http/endpoints/openai/realtime_chunker_test.go

@@ -0,0 +1,103 @@
+package openai
+
+import (
+	"strings"
+	"unicode/utf8"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+// clauseChunker splits streamed LLM content into speakable clauses in a
+// script-aware way: UAX#29 sentences (Latin .!? and CJK 。！？), CJK clause
+// punctuation, and Thai/Lao spaces — never whitespace-splitting CJK.
+var _ = Describe("clauseChunker", func() {
+	Context("Latin sentences", func() {
+		It("emits a sentence only once following content confirms it is complete", func() {
+			c := newClauseChunker(12, 200)
+			Expect(c.push("Hello world. How are you?")).To(Equal([]string{"Hello world."}))
+			// The trailing sentence is held until flush (the next token might extend it).
+			Expect(c.flush()).To(Equal([]string{"How are you?"}))
+		})
+
+		It("assembles a sentence across many small tokens", func() {
+			c := newClauseChunker(12, 200)
+			var got []string
+			for _, tok := range []string{"Hello", " world.", " How", " are", " you?"} {
+				got = append(got, c.push(tok)...)
+			}
+			got = append(got, c.flush()...)
+			Expect(got).To(Equal([]string{"Hello world.", "How are you?"}))
+		})
+
+		It("does not split decimals or abbreviations (UAX#29 SB6)", func() {
+			c := newClauseChunker(12, 200)
+			got := c.push("Pi is 3.14 and e is 2.72. Done")
+			Expect(got).To(Equal([]string{"Pi is 3.14 and e is 2.72."}))
+			Expect(c.flush()).To(Equal([]string{"Done"}))
+		})
+	})
+
+	Context("CJK (no whitespace)", func() {
+		It("splits Chinese on the ideographic full stop", func() {
+			c := newClauseChunker(12, 200)
+			Expect(c.push("你好世界。今天天气很好。")).To(Equal([]string{"你好世界。"}))
+			Expect(c.flush()).To(Equal([]string{"今天天气很好。"}))
+		})
+
+		It("splits Japanese on the ideographic full stop", func() {
+			c := newClauseChunker(12, 200)
+			Expect(c.push("こんにちは。元気ですか。")).To(Equal([]string{"こんにちは。"}))
+			Expect(c.flush()).To(Equal([]string{"元気ですか。"}))
+		})
+
+		It("splits on CJK clause punctuation for lower latency", func() {
+			c := newClauseChunker(2, 200) // small min so short test clauses cut
+			Expect(c.push("你好，世界。再见")).To(Equal([]string{"你好，", "世界。"}))
+			Expect(c.flush()).To(Equal([]string{"再见"}))
+		})
+	})
+
+	Context("Thai (spaces mark clauses, not words)", func() {
+		It("splits a Thai run on the inter-clause space", func() {
+			c := newClauseChunker(2, 200)
+			Expect(c.push("สวัสดีครับ กินข้าวไหม")).To(Equal([]string{"สวัสดีครับ"}))
+			Expect(c.flush()).To(Equal([]string{"กินข้าวไหม"}))
+		})
+
+		It("never shatters a space-less Thai run into characters", func() {
+			c := newClauseChunker(2, 200)
+			Expect(c.push("สวัสดีครับ")).To(BeEmpty()) // held, no boundary
+			Expect(c.flush()).To(Equal([]string{"สวัสดีครับ"}))
+		})
+	})
+
+	Context("length cap (UAX#14 fallback)", func() {
+		It("force-breaks an over-long punctuation-less CJK run at legal points", func() {
+			c := newClauseChunker(4, 10) // maxRunes = 10
+			run := strings.Repeat("字", 25)
+			got := c.push(run)
+			got = append(got, c.flush()...)
+			total := 0
+			for _, seg := range got {
+				n := utf8.RuneCountInString(seg)
+				Expect(n).To(BeNumerically("<=", 10)) // never exceeds the cap
+				total += n
+			}
+			Expect(total).To(Equal(25))                       // nothing dropped
+			Expect(len(got)).To(BeNumerically(">=", 3))       // 10 + 10 + 5
+		})
+	})
+
+	Context("buffer lifecycle", func() {
+		It("flush clears the buffer so the chunker is reusable", func() {
+			c := newClauseChunker(12, 200)
+			// "First one." is confirmed by the following "Second", so push drains it;
+			// only the unterminated tail remains for flush.
+			Expect(c.push("First one. Second")).To(Equal([]string{"First one."}))
+			Expect(c.flush()).To(Equal([]string{"Second"}))
+			Expect(c.flush()).To(BeEmpty())
+			Expect(c.push("Again. More")).To(Equal([]string{"Again."}))
+		})
+	})
+})

core/http/endpoints/openai/realtime_doubles_test.go

@@ -0,0 +1,138 @@
+package openai
+
+import (
+	"context"
+	"strings"
+
+	"github.com/mudler/LocalAI/core/backend"
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	"github.com/mudler/LocalAI/core/schema"
+	"github.com/mudler/LocalAI/pkg/grpc/proto"
+)
+
+// fakeTransport records the server events and audio sent to a realtime client
+// so streaming behaviour can be asserted without a real WebSocket/WebRTC peer.
+// It is not a *WebRTCTransport, so handler code takes the WebSocket path.
+type fakeTransport struct {
+	events []types.ServerEvent
+	audio  []fakeAudioChunk
+}
+
+type fakeAudioChunk struct {
+	pcm        []byte
+	sampleRate int
+}
+
+func (f *fakeTransport) SendEvent(e types.ServerEvent) error {
+	f.events = append(f.events, e)
+	return nil
+}
+
+func (f *fakeTransport) ReadEvent() ([]byte, error) { return nil, nil }
+
+func (f *fakeTransport) SendAudio(_ context.Context, pcm []byte, sampleRate int) error {
+	f.audio = append(f.audio, fakeAudioChunk{pcm: pcm, sampleRate: sampleRate})
+	return nil
+}
+
+func (f *fakeTransport) Close() error { return nil }
+
+// countEvents returns how many recorded events have the given type.
+func (f *fakeTransport) countEvents(et types.ServerEventType) int {
+	n := 0
+	for _, e := range f.events {
+		if e.ServerEventType() == et {
+			n++
+		}
+	}
+	return n
+}
+
+// transcriptDeltaText concatenates the Delta of every recorded transcript
+// delta event — i.e. the text streamed to the client as it is generated.
+func (f *fakeTransport) transcriptDeltaText() string {
+	var b strings.Builder
+	for _, e := range f.events {
+		if d, ok := e.(types.ResponseOutputAudioTranscriptDeltaEvent); ok {
+			b.WriteString(d.Delta)
+		}
+	}
+	return b.String()
+}
+
+// fakeModel is a configurable Model double. TTSStream replays ttsStreamChunks
+// and TranscribeStream replays transcribeDeltas, so the handler's streaming
+// paths can be driven deterministically.
+type fakeModel struct {
+	cfg *config.ModelConfig
+
+	ttsFile         string
+	ttsStreamChunks [][]byte
+	ttsStreamRate   int
+	ttsStreamErr    error
+
+	transcribeDeltas []string
+	transcribeFinal  *schema.TranscriptionResult
+
+	// Predict streaming: predictTokens are replayed through the token callback
+	// (simulating streamed LLM output); predictResp/predictErr are returned by
+	// the deferred predict function. predictChunkDeltas, when set, are delivered
+	// per-token via TokenUsage.ChatDeltas to exercise the autoparser path.
+	predictTokens      []string
+	predictChunkDeltas [][]*proto.ChatDelta
+	predictResp        backend.LLMResponse
+	predictErr         error
+}
+
+func (m *fakeModel) VAD(context.Context, *schema.VADRequest) (*schema.VADResponse, error) {
+	return nil, nil
+}
+
+func (m *fakeModel) Transcribe(context.Context, string, string, bool, bool, string) (*schema.TranscriptionResult, error) {
+	return m.transcribeFinal, nil
+}
+
+func (m *fakeModel) Predict(_ context.Context, _ schema.Messages, _, _, _ []string, cb func(string, backend.TokenUsage) bool, _ []types.ToolUnion, _ *types.ToolChoiceUnion, _, _ *int, _ map[string]float64) (func() (backend.LLMResponse, error), error) {
+	if m.predictErr != nil {
+		return nil, m.predictErr
+	}
+	return func() (backend.LLMResponse, error) {
+		for i, tok := range m.predictTokens {
+			if cb == nil {
+				continue
+			}
+			usage := backend.TokenUsage{}
+			if i < len(m.predictChunkDeltas) {
+				usage.ChatDeltas = m.predictChunkDeltas[i]
+			}
+			cb(tok, usage)
+		}
+		return m.predictResp, nil
+	}, nil
+}
+
+func (m *fakeModel) TTS(context.Context, string, string, string) (string, *proto.Result, error) {
+	return m.ttsFile, &proto.Result{Success: true}, nil
+}
+
+func (m *fakeModel) TTSStream(_ context.Context, _, _, _ string, onAudio func(pcm []byte, sampleRate int) error) error {
+	if m.ttsStreamErr != nil {
+		return m.ttsStreamErr
+	}
+	for _, c := range m.ttsStreamChunks {
+		if err := onAudio(c, m.ttsStreamRate); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func (m *fakeModel) TranscribeStream(_ context.Context, _, _ string, _, _ bool, _ string, onDelta func(text string)) (*schema.TranscriptionResult, error) {
+	for _, d := range m.transcribeDeltas {
+		onDelta(d)
+	}
+	return m.transcribeFinal, nil
+}
+
+func (m *fakeModel) PredictConfig() *config.ModelConfig { return m.cfg }

core/http/endpoints/openai/realtime_model.go

@@ -3,6 +3,7 @@ package openai
 import (
 	"context"
 	"crypto/rand"
+	"encoding/binary"
 	"encoding/hex"
 	"encoding/json"
 	"fmt"
@@ -87,6 +88,14 @@ func (m *transcriptOnlyModel) TTS(ctx context.Context, text, voice, language str
 	return "", nil, fmt.Errorf("TTS not supported in transcript-only mode")
 }
 
+func (m *transcriptOnlyModel) TTSStream(ctx context.Context, text, voice, language string, onAudio func(pcm []byte, sampleRate int) error) error {
+	return fmt.Errorf("TTS not supported in transcript-only mode")
+}
+
+func (m *transcriptOnlyModel) TranscribeStream(ctx context.Context, audio, language string, translate, diarize bool, prompt string, onDelta func(text string)) (*schema.TranscriptionResult, error) {
+	return transcribeStream(ctx, m.modelLoader, *m.TranscriptionConfig, m.appConfig, audio, language, translate, diarize, prompt, onDelta)
+}
+
 func (m *transcriptOnlyModel) PredictConfig() *config.ModelConfig {
 	return nil
 }
@@ -321,10 +330,75 @@ func (m *wrappedModel) TTS(ctx context.Context, text, voice, language string) (s
 	return backend.ModelTTS(ctx, text, voice, language, "", nil, m.modelLoader, m.appConfig, *m.TTSConfig)
 }
 
+func (m *wrappedModel) TTSStream(ctx context.Context, text, voice, language string, onAudio func(pcm []byte, sampleRate int) error) error {
+	return ttsStream(ctx, m.modelLoader, m.appConfig, *m.TTSConfig, text, voice, language, onAudio)
+}
+
+func (m *wrappedModel) TranscribeStream(ctx context.Context, audio, language string, translate, diarize bool, prompt string, onDelta func(text string)) (*schema.TranscriptionResult, error) {
+	return transcribeStream(ctx, m.modelLoader, *m.TranscriptionConfig, m.appConfig, audio, language, translate, diarize, prompt, onDelta)
+}
+
 func (m *wrappedModel) PredictConfig() *config.ModelConfig {
 	return m.LLMConfig
 }
 
+// wavStreamHeaderBytes is the size of the WAV header that backend.ModelTTSStream
+// emits as its first audio callback; the sample rate lives at byte offset 24.
+const wavStreamHeaderBytes = 44
+
+// ttsStream adapts backend.ModelTTSStream (which emits a WAV stream: a 44-byte
+// header carrying the sample rate, then raw PCM) to the realtime onAudio
+// callback, which wants raw PCM plus the sample rate. The header is buffered
+// until complete, the sample rate is read from it, and subsequent bytes are
+// forwarded as PCM.
+func ttsStream(ctx context.Context, ml *model.ModelLoader, appConfig *config.ApplicationConfig, ttsConfig config.ModelConfig, text, voice, language string, onAudio func(pcm []byte, sampleRate int) error) error {
+	var header []byte
+	headerDone := false
+	sampleRate := 0
+	return backend.ModelTTSStream(ctx, text, voice, language, "", nil, ml, appConfig, ttsConfig, func(b []byte) error {
+		if headerDone {
+			if len(b) == 0 {
+				return nil
+			}
+			return onAudio(b, sampleRate)
+		}
+		header = append(header, b...)
+		if len(header) < wavStreamHeaderBytes {
+			return nil
+		}
+		sampleRate = int(binary.LittleEndian.Uint32(header[24:28]))
+		headerDone = true
+		if len(header) > wavStreamHeaderBytes {
+			return onAudio(header[wavStreamHeaderBytes:], sampleRate)
+		}
+		return nil
+	})
+}
+
+// transcribeStream adapts backend.ModelTranscriptionStream to the realtime
+// onDelta callback, returning the final aggregated transcription result.
+func transcribeStream(ctx context.Context, ml *model.ModelLoader, transcriptionConfig config.ModelConfig, appConfig *config.ApplicationConfig, audio, language string, translate, diarize bool, prompt string, onDelta func(text string)) (*schema.TranscriptionResult, error) {
+	var final *schema.TranscriptionResult
+	err := backend.ModelTranscriptionStream(ctx, backend.TranscriptionRequest{
+		Audio:     audio,
+		Language:  language,
+		Translate: translate,
+		Diarize:   diarize,
+		Prompt:    prompt,
+	}, ml, transcriptionConfig, appConfig, func(chunk backend.TranscriptionStreamChunk) {
+		if chunk.Delta != "" {
+			onDelta(chunk.Delta)
+		}
+		if chunk.Final != nil {
+			final = chunk.Final
+		}
+	})
+	if err != nil {
+		return nil, err
+	}
+	return final, nil
+}
+
 func newTranscriptionOnlyModel(pipeline *config.Pipeline, cl *config.ModelConfigLoader, ml *model.ModelLoader, appConfig *config.ApplicationConfig) (Model, *config.ModelConfig, error) {
 	cfgVAD, err := cl.LoadModelConfigFileByName(pipeline.VAD, ml.ModelPath)
 	if err != nil {
@@ -454,8 +528,10 @@ func newModel(pipeline *config.Pipeline, cl *config.ModelConfigLoader, ml *model
 		return nil, fmt.Errorf("failed to validate config: %w", err)
 	}
 
-	// Let the pipeline set the LLM's reasoning effort (cfgLLM is a per-session copy).
+	// Let the pipeline set the LLM's reasoning effort and force thinking off
+	// (cfgLLM is a per-session copy). disable_thinking applies after the effort.
 	applyPipelineReasoning(cfgLLM, *pipeline)
+	applyPipelineThinking(cfgLLM, *pipeline)
 
 	cfgTTS, err := cl.LoadModelConfigFileByName(pipeline.TTS, ml.ModelPath)
 	if err != nil {

core/http/endpoints/openai/realtime_speech.go

@@ -0,0 +1,102 @@
+package openai
+
+import (
+	"context"
+	"encoding/base64"
+	"fmt"
+	"os"
+	"path/filepath"
+
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	laudio "github.com/mudler/LocalAI/pkg/audio"
+	"github.com/mudler/LocalAI/pkg/sound"
+)
+
+// emitSpeech synthesizes text and sends the audio to the client. When the
+// pipeline opts into TTS streaming it forwards each PCM chunk as its own
+// response.output_audio.delta as soon as the backend produces it; otherwise it
+// synthesizes the whole utterance and sends it as a single delta.
+//
+// It deliberately does NOT emit transcript or audio-done events: the caller owns
+// those so a streamed reply can be split into several spoken segments that share
+// one response/item.
+//
+// It returns the PCM audio (at the session output rate) accumulated across all
+// chunks, which the caller base64-encodes onto the conversation item. For WebRTC
+// the audio goes over the RTP track instead, so the returned slice is empty.
+func emitSpeech(ctx context.Context, t Transport, session *Session, responseID, itemID, text string) ([]byte, error) {
+	if text == "" {
+		return nil, nil
+	}
+
+	_, isWebRTC := t.(*WebRTCTransport)
+
+	var wsAudio []byte // PCM at the session output rate, accumulated for the item record
+
+	// sendChunk hands one PCM buffer to the transport: WebRTC consumes the raw
+	// PCM directly (it resamples internally); WebSocket gets base64 PCM at the
+	// session output rate via a JSON delta event.
+	sendChunk := func(pcm []byte, sampleRate int) error {
+		if len(pcm) == 0 {
+			return nil
+		}
+		if err := t.SendAudio(ctx, pcm, sampleRate); err != nil {
+			return err
+		}
+		if isWebRTC {
+			return nil
+		}
+		wsPCM := pcm
+		if sampleRate != 0 && sampleRate != session.OutputSampleRate {
+			samples := sound.BytesToInt16sLE(pcm)
+			resampled := sound.ResampleInt16(samples, sampleRate, session.OutputSampleRate)
+			wsPCM = sound.Int16toBytesLE(resampled)
+		}
+		wsAudio = append(wsAudio, wsPCM...)
+		return t.SendEvent(types.ResponseOutputAudioDeltaEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			ItemID:          itemID,
+			OutputIndex:     0,
+			ContentIndex:    0,
+			Delta:           base64.StdEncoding.EncodeToString(wsPCM),
+		})
+	}
+
+	language := ""
+	if session.InputAudioTranscription != nil {
+		language = session.InputAudioTranscription.Language
+	}
+
+	if session.ModelConfig != nil && session.ModelConfig.Pipeline.StreamTTS() {
+		if err := session.ModelInterface.TTSStream(ctx, text, session.Voice, language, sendChunk); err != nil {
+			return nil, err
+		}
+		return wsAudio, nil
+	}
+
+	// Unary fallback: synthesize the whole utterance to a file, then emit once.
+	audioFilePath, res, err := session.ModelInterface.TTS(ctx, text, session.Voice, language)
+	if err != nil {
+		return nil, err
+	}
+	if res != nil && !res.Success {
+		return nil, fmt.Errorf("tts generation failed: %s", res.Message)
+	}
+	defer func() { _ = os.Remove(audioFilePath) }()
+
+	// filepath.Clean normalizes the backend-produced temp path before reading
+	// (also keeps gosec G304 quiet — the path is backend-controlled, not user input).
+	audioBytes, err := os.ReadFile(filepath.Clean(audioFilePath))
+	if err != nil {
+		return nil, fmt.Errorf("read tts audio: %w", err)
+	}
+	pcm, sampleRate := laudio.ParseWAV(audioBytes)
+	if sampleRate == 0 {
+		sampleRate = session.OutputSampleRate
+	}
+	if err := sendChunk(pcm, sampleRate); err != nil {
+		return nil, err
+	}
+	return wsAudio, nil
+}

core/http/endpoints/openai/realtime_speech_test.go

@@ -0,0 +1,70 @@
+package openai
+
+import (
+	"context"
+	"os"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	laudio "github.com/mudler/LocalAI/pkg/audio"
+)
+
+// emitSpeech synthesizes a piece of text and forwards the audio to the client,
+// streaming a delta per TTS chunk when the pipeline opts in, or sending the
+// whole utterance as one delta otherwise.
+var _ = Describe("emitSpeech", func() {
+	ttsOn := true
+
+	streamingSession := func(m Model) *Session {
+		return &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   m,
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{TTS: &ttsOn}},
+			},
+		}
+	}
+
+	It("streams one output_audio.delta per TTS chunk when streaming is enabled", func() {
+		m := &fakeModel{
+			ttsStreamChunks: [][]byte{{1, 2}, {3, 4}, {5, 6}},
+			ttsStreamRate:   24000,
+		}
+		t := &fakeTransport{}
+
+		audio, err := emitSpeech(context.Background(), t, streamingSession(m), "resp1", "item1", "Hello there.")
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioDelta)).To(Equal(3))
+		// The returned audio is all chunks concatenated (session output rate).
+		Expect(audio).To(Equal([]byte{1, 2, 3, 4, 5, 6}))
+	})
+
+	It("sends a single output_audio.delta in unary mode", func() {
+		// A minimal real WAV file for the unary TTS path to read + parse.
+		f, err := os.CreateTemp("", "emit-*.wav")
+		Expect(err).ToNot(HaveOccurred())
+		defer func() { _ = os.Remove(f.Name()) }()
+		pcm := make([]byte, 320) // 160 samples of silence
+		hdr := laudio.NewWAVHeader(uint32(len(pcm)))
+		Expect(hdr.Write(f)).To(Succeed())
+		_, err = f.Write(pcm)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(f.Close()).To(Succeed())
+
+		session := &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   &fakeModel{ttsFile: f.Name()},
+			ModelConfig:      &config.ModelConfig{}, // streaming off
+		}
+		t := &fakeTransport{}
+
+		_, err = emitSpeech(context.Background(), t, session, "resp1", "item1", "Hello there.")
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioDelta)).To(Equal(1))
+	})
+})

core/http/endpoints/openai/realtime_stream.go

@@ -0,0 +1,315 @@
+package openai
+
+import (
+	"context"
+	"encoding/base64"
+	"fmt"
+
+	"github.com/mudler/LocalAI/core/backend"
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	"github.com/mudler/LocalAI/core/schema"
+	"github.com/mudler/LocalAI/pkg/functions"
+	"github.com/mudler/LocalAI/pkg/reasoning"
+)
+
+// transcriptStreamer turns streamed LLM tokens into the assistant's spoken
+// transcript: it strips reasoning incrementally and sends one
+// response.output_audio_transcript.delta per content fragment. It does NOT
+// synthesize audio — the caller buffers the full message and synthesizes it
+// once (streaming the audio chunks when the TTS backend supports TTSStream),
+// which works uniformly for streaming and non-streaming TTS and for languages
+// without sentence or word boundaries.
+type transcriptStreamer struct {
+	ctx        context.Context
+	t          Transport
+	responseID string
+	itemID     string
+	extractor  *reasoning.ReasoningExtractor
+
+	// announce, if set, is invoked once just before the first transcript delta.
+	// It lets the caller create the assistant item lazily, so a content-less
+	// tool-call turn never emits a spurious empty assistant item.
+	announce  func()
+	announced bool
+}
+
+func newTranscriptStreamer(ctx context.Context, t Transport, responseID, itemID, thinkingStartToken string, reasoningCfg reasoning.Config) *transcriptStreamer {
+	return &transcriptStreamer{
+		ctx:        ctx,
+		t:          t,
+		responseID: responseID,
+		itemID:     itemID,
+		extractor:  reasoning.NewReasoningExtractor(thinkingStartToken, spokenReasoningConfig(reasoningCfg)),
+	}
+}
+
+// onToken handles one streamed unit of model output, sending a transcript delta
+// for the new content (reasoning stripped) and returning that content delta so
+// the caller can also feed it to the clause chunker. For plain-content models
+// the unit is the raw text token; for autoparser tool turns the backend clears
+// the text and delivers content via ChatDeltas, so the caller passes that
+// content here. Returns "" when the token produced no new spoken content.
+func (s *transcriptStreamer) onToken(token string) string {
+	_, content := s.extractor.ProcessToken(token)
+	if content == "" {
+		return ""
+	}
+	if !s.announced {
+		s.announced = true
+		if s.announce != nil {
+			s.announce()
+		}
+	}
+	_ = s.t.SendEvent(types.ResponseOutputAudioTranscriptDeltaEvent{
+		ServerEventBase: types.ServerEventBase{},
+		ResponseID:      s.responseID,
+		ItemID:          s.itemID,
+		OutputIndex:     0,
+		ContentIndex:    0,
+		Delta:           content,
+	})
+	return content
+}
+
+// content returns the full transcript so far with reasoning stripped.
+func (s *transcriptStreamer) content() string {
+	return s.extractor.CleanedContent()
+}
+
+// streamLLMResponse drives a streamed realtime reply. It streams the assistant
+// transcript as the LLM generates, then synthesizes the whole buffered message
+// once (streaming the audio chunks when the TTS backend supports it, otherwise a
+// single unary delta). Tool calls parsed from the autoparser ChatDeltas are
+// emitted after the spoken content. The assistant content item is created lazily
+// on the first content delta, so a content-less tool-call turn emits only the
+// tool calls. It returns true when it has fully handled the response so the
+// caller can return; callers must only invoke it for an audio modality, and with
+// tools only when the model uses its tokenizer template (see triggerResponseAtTurn).
+func streamLLMResponse(ctx context.Context, session *Session, conv *Conversation, t Transport, responseID string, history schema.Messages, images []string, llmCfg *config.ModelConfig, tools []types.ToolUnion, toolChoice *types.ToolChoiceUnion, toolTurn int) bool {
+	itemID := generateItemID()
+	item := types.MessageItemUnion{
+		Assistant: &types.MessageItemAssistant{
+			ID:      itemID,
+			Status:  types.ItemStatusInProgress,
+			Content: []types.MessageContentOutput{{Type: types.MessageContentTypeOutputAudio}},
+		},
+	}
+
+	// announce creates the assistant content item lazily, just before the first
+	// transcript delta — a tool-only turn never produces content, so it stays out
+	// of the conversation and the client sees only the tool calls.
+	announced := false
+	announce := func() {
+		announced = true
+		conv.Lock.Lock()
+		conv.Items = append(conv.Items, &item)
+		conv.Lock.Unlock()
+		sendEvent(t, types.ResponseOutputItemAddedEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			OutputIndex:     0,
+			Item:            item,
+		})
+		sendEvent(t, types.ResponseContentPartAddedEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			ItemID:          itemID,
+			OutputIndex:     0,
+			ContentIndex:    0,
+			Part:            item.Assistant.Content[0],
+		})
+	}
+
+	cancel := func() {
+		if announced {
+			conv.Lock.Lock()
+			for i := len(conv.Items) - 1; i >= 0; i-- {
+				if conv.Items[i].Assistant != nil && conv.Items[i].Assistant.ID == itemID {
+					conv.Items = append(conv.Items[:i], conv.Items[i+1:]...)
+					break
+				}
+			}
+			conv.Lock.Unlock()
+		}
+		sendEvent(t, types.ResponseDoneEvent{
+			ServerEventBase: types.ServerEventBase{},
+			Response:        types.Response{ID: responseID, Object: "realtime.response", Status: types.ResponseStatusCancelled},
+		})
+	}
+
+	var template string
+	if llmCfg.TemplateConfig.UseTokenizerTemplate {
+		template = llmCfg.GetModelTemplate()
+	} else {
+		template = llmCfg.TemplateConfig.Chat
+	}
+	thinkingStartToken := reasoning.DetectThinkingStartToken(template, &llmCfg.ReasoningConfig)
+
+	// The autoparser (tokenizer-template path) already delivers reasoning-free
+	// content. Prefilling the thinking start token here would re-tag that clean
+	// content as an unclosed reasoning block, leaving CleanedContent() empty —
+	// no spoken reply, no TTS. Disable the prefill; closed tag pairs are still
+	// stripped (PEG-fallback case, #9985).
+	reasoningCfg := llmCfg.ReasoningConfig
+	if llmCfg.TemplateConfig.UseTokenizerTemplate {
+		disablePrefill := true
+		reasoningCfg.DisableReasoningTagPrefill = &disablePrefill
+	}
+
+	streamer := newTranscriptStreamer(ctx, t, responseID, itemID, thinkingStartToken, reasoningCfg)
+	streamer.announce = announce
+
+	// Clause chunking (opt-in): synthesize each clause as soon as it completes
+	// instead of buffering the whole reply. streamedAudio accumulates the PCM
+	// across clauses for the conversation item record; ttsErr captures the first
+	// synthesis failure so the token callback can stop the prediction. emitSpeech
+	// runs synchronously here — the LLM keeps generating into the gRPC stream
+	// while a clause is synthesized, so audio still starts mid-generation.
+	var chunker *clauseChunker
+	if session.ModelConfig != nil && session.ModelConfig.Pipeline.ChunkClauses() {
+		chunker = newClauseChunker(defaultClauseMinRunes, defaultClauseMaxRunes)
+	}
+	var streamedAudio []byte
+	var ttsErr error
+	speakClause := func(clause string) error {
+		a, err := emitSpeech(ctx, t, session, responseID, itemID, clause)
+		if err != nil {
+			return err
+		}
+		streamedAudio = append(streamedAudio, a...)
+		return nil
+	}
+
+	// fail reports a mid-stream failure. A cancelled context means the client
+	// interrupted (barge-in), so roll the turn back instead of erroring.
+	fail := func(code, msg string, err error) bool {
+		if ctx.Err() != nil {
+			cancel()
+		} else {
+			sendError(t, code, fmt.Sprintf("%s: %v", msg, err), "", itemID)
+		}
+		return true
+	}
+
+	cb := func(token string, usage backend.TokenUsage) bool {
+		if ctx.Err() != nil {
+			return false
+		}
+		// Plain-content models stream text via the token; autoparser tool turns
+		// clear the text and deliver content via ChatDeltas, so prefer the latter
+		// when present. Either way only content reaches the transcript — tool-call
+		// deltas are parsed from the final response below.
+		text := token
+		if len(usage.ChatDeltas) > 0 {
+			text = functions.ContentFromChatDeltas(usage.ChatDeltas)
+		}
+		delta := streamer.onToken(text)
+		if chunker != nil && delta != "" {
+			for _, clause := range chunker.push(delta) {
+				if ttsErr = speakClause(clause); ttsErr != nil {
+					return false // stop the prediction; reported after predFunc returns
+				}
+			}
+		}
+		return true
+	}
+
+	predFunc, err := session.ModelInterface.Predict(ctx, history, images, nil, nil, cb, tools, toolChoice, nil, nil, nil)
+	if err != nil {
+		sendError(t, "inference_failed", fmt.Sprintf("backend error: %v", err), "", itemID)
+		return true
+	}
+	pred, err := predFunc()
+	// A clause synthesis failed mid-stream (the callback stopped the prediction);
+	// report it as a TTS error rather than a prediction error.
+	if ttsErr != nil {
+		return fail("tts_error", "TTS generation failed", ttsErr)
+	}
+	if err != nil {
+		return fail("prediction_failed", "backend error", err)
+	}
+	if ctx.Err() != nil {
+		cancel()
+		return true
+	}
+
+	content := streamer.content()
+	toolCalls := functions.ToolCallsFromChatDeltas(pred.ChatDeltas)
+
+	// Finalize the spoken content item only when the turn produced content. A
+	// tool-only turn skips this entirely (no empty assistant item).
+	if content != "" {
+		if !announced {
+			announce()
+		}
+
+		// Synthesize the audio. With clause chunking the completed clauses were
+		// already spoken inside the token callback; flush the trailing clause(s)
+		// the segmenter was still holding. Otherwise buffer the whole message and
+		// synthesize it once. emitSpeech streams the audio chunks when the TTS
+		// backend supports TTSStream, otherwise it sends a single unary delta.
+		var audio []byte
+		if chunker != nil {
+			for _, clause := range chunker.flush() {
+				if ttsErr = speakClause(clause); ttsErr != nil {
+					break
+				}
+			}
+			audio = streamedAudio
+		} else {
+			audio, ttsErr = emitSpeech(ctx, t, session, responseID, itemID, content)
+		}
+		if ttsErr != nil {
+			return fail("tts_error", "TTS generation failed", ttsErr)
+		}
+
+		_, isWebRTC := t.(*WebRTCTransport)
+
+		sendEvent(t, types.ResponseOutputAudioTranscriptDoneEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			ItemID:          itemID,
+			OutputIndex:     0,
+			ContentIndex:    0,
+			Transcript:      content,
+		})
+		if !isWebRTC {
+			sendEvent(t, types.ResponseOutputAudioDoneEvent{
+				ServerEventBase: types.ServerEventBase{},
+				ResponseID:      responseID,
+				ItemID:          itemID,
+				OutputIndex:     0,
+				ContentIndex:    0,
+			})
+		}
+
+		conv.Lock.Lock()
+		item.Assistant.Status = types.ItemStatusCompleted
+		item.Assistant.Content[0].Transcript = content
+		if !isWebRTC {
+			item.Assistant.Content[0].Audio = base64.StdEncoding.EncodeToString(audio)
+		}
+		conv.Lock.Unlock()
+
+		sendEvent(t, types.ResponseContentPartDoneEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			ItemID:          itemID,
+			OutputIndex:     0,
+			ContentIndex:    0,
+			Part:            item.Assistant.Content[0],
+		})
+		sendEvent(t, types.ResponseOutputItemDoneEvent{
+			ServerEventBase: types.ServerEventBase{},
+			ResponseID:      responseID,
+			OutputIndex:     0,
+			Item:            item,
+		})
+	}
+
+	// Emit any tool calls, the terminal response.done, and (for server-side
+	// assistant tools) the follow-up turn — shared with the buffered path.
+	emitToolCallItems(ctx, session, conv, t, responseID, toolCalls, content != "", toolTurn)
+	return true
+}

core/http/endpoints/openai/realtime_stream_test.go

@@ -0,0 +1,213 @@
+package openai
+
+import (
+	"context"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/backend"
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	"github.com/mudler/LocalAI/pkg/grpc/proto"
+	"github.com/mudler/LocalAI/pkg/reasoning"
+)
+
+// transcriptStreamer turns streamed LLM tokens into incremental transcript
+// deltas, stripping reasoning. Audio is synthesized once from the full message
+// by the caller, so there is no per-sentence segmentation.
+var _ = Describe("transcriptStreamer", func() {
+	It("emits one transcript delta per content token", func() {
+		t := &fakeTransport{}
+		s := newTranscriptStreamer(context.Background(), t, "resp1", "item1", "", reasoning.Config{})
+
+		for _, tok := range []string{"Hello", " world.", " Bye"} {
+			s.onToken(tok)
+		}
+
+		Expect(s.content()).To(Equal("Hello world. Bye"))
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioTranscriptDelta)).To(Equal(3))
+		Expect(t.transcriptDeltaText()).To(Equal("Hello world. Bye"))
+	})
+
+	It("strips leaked reasoning even when reasoning is disabled (disable_thinking safety net)", func() {
+		// disable_thinking maps to DisableReasoning=true (enable_thinking=false to
+		// the backend). If the model emits thinking anyway, the transcript must
+		// still not leak it: stripping always runs for spoken output.
+		disable := true
+		t := &fakeTransport{}
+		s := newTranscriptStreamer(context.Background(), t, "resp1", "item1", "",
+			reasoning.Config{DisableReasoning: &disable})
+
+		s.onToken("<think>secret plan</think>")
+		s.onToken("The answer is 42.")
+
+		Expect(s.content()).To(Equal("The answer is 42."))
+		Expect(s.content()).ToNot(ContainSubstring("secret plan"))
+		Expect(t.transcriptDeltaText()).ToNot(ContainSubstring("secret plan"))
+	})
+
+	It("does not swallow autoparser content when the template has a thinking start token (tokenizer-template path)", func() {
+		// Regression: with tag prefill on, the detected <think> token is
+		// prepended to the autoparser's already-clean content, swallowing the
+		// whole reply (empty transcript → no TTS). streamLLMResponse disables
+		// the prefill for the tokenizer-template path.
+		disablePrefill := true
+		t := &fakeTransport{}
+		s := newTranscriptStreamer(context.Background(), t, "resp1", "item1", "<think>",
+			reasoning.Config{DisableReasoningTagPrefill: &disablePrefill})
+
+		s.onToken("Hello")
+		s.onToken(" there.")
+
+		Expect(s.content()).To(Equal("Hello there."))
+		Expect(t.transcriptDeltaText()).To(Equal("Hello there."))
+	})
+
+	It("still strips embedded closed reasoning tags with prefill disabled (PEG-fallback safety, #9985)", func() {
+		// Disabling prefill must not stop stripping closed <think>…</think>
+		// pairs the PEG fallback can leave in autoparser content.
+		disablePrefill := true
+		t := &fakeTransport{}
+		s := newTranscriptStreamer(context.Background(), t, "resp1", "item1", "<think>",
+			reasoning.Config{DisableReasoningTagPrefill: &disablePrefill})
+
+		s.onToken("<think>secret</think>")
+		s.onToken("The answer is 42.")
+
+		Expect(s.content()).To(Equal("The answer is 42."))
+		Expect(t.transcriptDeltaText()).ToNot(ContainSubstring("secret"))
+	})
+})
+
+// streamLLMResponse drives a full streamed realtime turn: live transcript
+// deltas while the LLM generates, then the whole message is synthesized once.
+var _ = Describe("streamLLMResponse", func() {
+	It("streams transcript deltas then synthesizes the whole message once", func() {
+		on := true
+		m := &fakeModel{
+			predictTokens:   []string{"Hello", " world.", " How are you?"},
+			predictResp:     backend.LLMResponse{Response: "Hello world. How are you?"},
+			ttsStreamChunks: [][]byte{{9}},
+			ttsStreamRate:   24000,
+		}
+		session := &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   m,
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{LLM: &on, TTS: &on}},
+			},
+		}
+		conv := &Conversation{}
+		t := &fakeTransport{}
+		llmCfg := &config.ModelConfig{}
+
+		handled := streamLLMResponse(context.Background(), session, conv, t, "resp1", nil, nil, llmCfg, nil, nil, 0)
+
+		Expect(handled).To(BeTrue())
+		// One live transcript delta per streamed token.
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioTranscriptDelta)).To(Equal(3))
+		// The whole message is synthesized ONCE (not per sentence): a single
+		// emitSpeech replays the one TTS stream chunk.
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioDelta)).To(Equal(1))
+		Expect(t.transcriptDeltaText()).To(Equal("Hello world. How are you?"))
+	})
+
+	It("synthesizes each clause as it completes when clause chunking is enabled", func() {
+		on := true
+		m := &fakeModel{
+			predictTokens:   []string{"Hello world.", " How are you?"},
+			predictResp:     backend.LLMResponse{Response: "Hello world. How are you?"},
+			ttsStreamChunks: [][]byte{{9}},
+			ttsStreamRate:   24000,
+		}
+		session := &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   m,
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{LLM: &on, TTS: &on, ClauseChunking: &on}},
+			},
+		}
+		conv := &Conversation{}
+		t := &fakeTransport{}
+		llmCfg := &config.ModelConfig{}
+
+		handled := streamLLMResponse(context.Background(), session, conv, t, "resp1", nil, nil, llmCfg, nil, nil, 0)
+
+		Expect(handled).To(BeTrue())
+		// Two clauses ("Hello world." mid-stream, "How are you?" on flush) → two
+		// emitSpeech calls → two audio deltas, vs one for whole-message buffering.
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioDelta)).To(Equal(2))
+		// The full transcript still streams verbatim.
+		Expect(t.transcriptDeltaText()).To(Equal("Hello world. How are you?"))
+		// Exactly one terminal response.done.
+		Expect(t.countEvents(types.ServerEventTypeResponseDone)).To(Equal(1))
+	})
+
+	It("streams content deltas and emits tool-call items (autoparser tool turn)", func() {
+		on := true
+		// Autoparser path: reply.Message is empty; content + tool calls arrive via
+		// ChatDeltas. Chunk 1 carries content, chunk 2 carries the tool call.
+		contentDelta := []*proto.ChatDelta{{Content: "Let me check."}}
+		toolDelta := []*proto.ChatDelta{{ToolCalls: []*proto.ToolCallDelta{{Index: 0, Name: "get_weather", Arguments: `{"city":"Paris"}`}}}}
+		m := &fakeModel{
+			predictTokens:      []string{"", ""},
+			predictChunkDeltas: [][]*proto.ChatDelta{contentDelta, toolDelta},
+			predictResp:        backend.LLMResponse{ChatDeltas: append(append([]*proto.ChatDelta{}, contentDelta...), toolDelta...)},
+			ttsStreamChunks:    [][]byte{{9}},
+			ttsStreamRate:      24000,
+		}
+		session := &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   m,
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{LLM: &on, TTS: &on}},
+			},
+		}
+		conv := &Conversation{}
+		t := &fakeTransport{}
+		llmCfg := &config.ModelConfig{}
+		llmCfg.TemplateConfig.UseTokenizerTemplate = true
+
+		handled := streamLLMResponse(context.Background(), session, conv, t, "resp1", nil, nil, llmCfg, nil, nil, 0)
+
+		Expect(handled).To(BeTrue())
+		// The spoken content was streamed live.
+		Expect(t.transcriptDeltaText()).To(Equal("Let me check."))
+		// The tool call is emitted as a function_call item.
+		Expect(t.countEvents(types.ServerEventTypeResponseFunctionCallArgumentsDone)).To(Equal(1))
+		// Exactly one terminal response.done.
+		Expect(t.countEvents(types.ServerEventTypeResponseDone)).To(Equal(1))
+	})
+
+	It("emits only tool-call items for a content-less tool turn (no empty assistant item)", func() {
+		on := true
+		toolDelta := []*proto.ChatDelta{{ToolCalls: []*proto.ToolCallDelta{{Index: 0, Name: "get_weather", Arguments: `{"city":"Rome"}`}}}}
+		m := &fakeModel{
+			predictTokens:      []string{""},
+			predictChunkDeltas: [][]*proto.ChatDelta{toolDelta},
+			predictResp:        backend.LLMResponse{ChatDeltas: toolDelta},
+		}
+		session := &Session{
+			OutputSampleRate: 24000,
+			ModelInterface:   m,
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{LLM: &on, TTS: &on}},
+			},
+		}
+		conv := &Conversation{}
+		t := &fakeTransport{}
+		llmCfg := &config.ModelConfig{}
+		llmCfg.TemplateConfig.UseTokenizerTemplate = true
+
+		handled := streamLLMResponse(context.Background(), session, conv, t, "resp1", nil, nil, llmCfg, nil, nil, 0)
+
+		Expect(handled).To(BeTrue())
+		// No content → no transcript deltas and no spurious assistant content item.
+		Expect(t.transcriptDeltaText()).To(Equal(""))
+		Expect(t.countEvents(types.ServerEventTypeResponseOutputAudioTranscriptDelta)).To(Equal(0))
+		// The tool call is still emitted.
+		Expect(t.countEvents(types.ServerEventTypeResponseFunctionCallArgumentsDone)).To(Equal(1))
+		Expect(t.countEvents(types.ServerEventTypeResponseDone)).To(Equal(1))
+	})
+})

core/http/endpoints/openai/realtime_thinking.go

@@ -0,0 +1,33 @@
+package openai
+
+import (
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/pkg/reasoning"
+)
+
+// applyPipelineThinking forces the LLM's reasoning/thinking off when the realtime
+// pipeline sets disable_thinking, mapping to the enable_thinking=false backend
+// metadata via ReasoningConfig.DisableReasoning. The LLM config passed in is the
+// per-session copy returned by the config loader, so this does not affect other
+// users of the same model. When the pipeline does not set disable_thinking the
+// LLM config is left untouched.
+func applyPipelineThinking(llm *config.ModelConfig, pipeline config.Pipeline) {
+	if llm == nil || !pipeline.ThinkingDisabled() {
+		return
+	}
+	disable := true
+	llm.ReasoningConfig.DisableReasoning = &disable
+}
+
+// spokenReasoningConfig adapts a model's reasoning config for stripping reasoning
+// OUT of realtime spoken output. ReasoningConfig.DisableReasoning is overloaded:
+// the backend reads it as the "enable_thinking=false" hint (which pipeline
+// disable_thinking sets via applyPipelineThinking), but the reasoning extractor
+// reads it as "skip stripping, assume there is no reasoning". Honouring the latter
+// when extracting for speech would leak raw <think>…</think> whenever the model
+// ignores the suppression hint. Spoken output must never contain reasoning, so we
+// always strip: clear DisableReasoning while keeping custom tokens/tag pairs.
+func spokenReasoningConfig(cfg reasoning.Config) reasoning.Config {
+	cfg.DisableReasoning = nil
+	return cfg
+}

core/http/endpoints/openai/realtime_thinking_test.go

@@ -0,0 +1,50 @@
+package openai
+
+import (
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/pkg/reasoning"
+)
+
+// applyPipelineThinking lets a realtime pipeline force the LLM's thinking off
+// (enable_thinking=false metadata) without editing the LLM model config.
+var _ = Describe("applyPipelineThinking", func() {
+	It("disables reasoning on the LLM config when the pipeline disables thinking", func() {
+		disable := true
+		llm := &config.ModelConfig{}
+		applyPipelineThinking(llm, config.Pipeline{DisableThinking: &disable})
+		Expect(llm.ReasoningConfig.DisableReasoning).ToNot(BeNil())
+		Expect(*llm.ReasoningConfig.DisableReasoning).To(BeTrue())
+	})
+
+	It("leaves the LLM config untouched when the pipeline does not set disable_thinking", func() {
+		llm := &config.ModelConfig{}
+		applyPipelineThinking(llm, config.Pipeline{})
+		Expect(llm.ReasoningConfig.DisableReasoning).To(BeNil())
+	})
+})
+
+// spokenReasoningConfig clears DisableReasoning so realtime spoken output always
+// strips reasoning, even though disable_thinking sets DisableReasoning=true on the
+// LLM config (which the backend reads as enable_thinking=false).
+var _ = Describe("spokenReasoningConfig", func() {
+	It("clears DisableReasoning so the extractor still strips leaked reasoning", func() {
+		disable := true
+		out := spokenReasoningConfig(reasoning.Config{DisableReasoning: &disable})
+		Expect(out.DisableReasoning).To(BeNil())
+	})
+
+	It("preserves the other reasoning settings", func() {
+		disable := true
+		out := spokenReasoningConfig(reasoning.Config{
+			DisableReasoning:    &disable,
+			ThinkingStartTokens: []string{"<reason>"},
+			TagPairs:            []reasoning.TagPair{{Start: "<reason>", End: "</reason>"}},
+		})
+		Expect(out.ThinkingStartTokens).To(Equal([]string{"<reason>"}))
+		Expect(out.TagPairs).To(HaveLen(1))
+		Expect(out.TagPairs[0].Start).To(Equal("<reason>"))
+	})
+})

core/http/endpoints/openai/realtime_transcription.go

@@ -0,0 +1,63 @@
+package openai
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+)
+
+// emitTranscription transcribes a committed utterance and emits the transcription
+// events for it, returning the final transcript text. With
+// pipeline.streaming.transcription enabled it streams each transcript fragment as
+// a conversation.item.input_audio_transcription.delta as the backend produces it,
+// then a completed event; otherwise it transcribes the whole utterance and emits
+// a single completed event. delta and completed events share itemID.
+func emitTranscription(ctx context.Context, t Transport, session *Session, itemID, audioPath string) (string, error) {
+	cfg := session.InputAudioTranscription
+
+	if session.ModelConfig != nil && session.ModelConfig.Pipeline.StreamTranscription() {
+		final, err := session.ModelInterface.TranscribeStream(ctx, audioPath, cfg.Language, false, false, cfg.Prompt, func(delta string) {
+			_ = t.SendEvent(types.ConversationItemInputAudioTranscriptionDeltaEvent{
+				ServerEventBase: types.ServerEventBase{EventID: "event_TODO"},
+				ItemID:          itemID,
+				ContentIndex:    0,
+				Delta:           delta,
+			})
+		})
+		if err != nil {
+			return "", err
+		}
+		transcript := ""
+		if final != nil {
+			transcript = final.Text
+		}
+		if err := t.SendEvent(types.ConversationItemInputAudioTranscriptionCompletedEvent{
+			ServerEventBase: types.ServerEventBase{EventID: "event_TODO"},
+			ItemID:          itemID,
+			ContentIndex:    0,
+			Transcript:      transcript,
+		}); err != nil {
+			return "", err
+		}
+		return transcript, nil
+	}
+
+	// Unary fallback: transcribe the whole utterance, emit one completed event.
+	tr, err := session.ModelInterface.Transcribe(ctx, audioPath, cfg.Language, false, false, cfg.Prompt)
+	if err != nil {
+		return "", err
+	}
+	if tr == nil {
+		return "", fmt.Errorf("transcribe result is nil")
+	}
+	if err := t.SendEvent(types.ConversationItemInputAudioTranscriptionCompletedEvent{
+		ServerEventBase: types.ServerEventBase{EventID: "event_TODO"},
+		ItemID:          itemID,
+		ContentIndex:    0,
+		Transcript:      tr.Text,
+	}); err != nil {
+		return "", err
+	}
+	return tr.Text, nil
+}

core/http/endpoints/openai/realtime_transcription_test.go

@@ -0,0 +1,54 @@
+package openai
+
+import (
+	"context"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/http/endpoints/openai/types"
+	"github.com/mudler/LocalAI/core/schema"
+)
+
+// emitTranscription transcribes a committed utterance, streaming transcript text
+// deltas when the pipeline opts in, and returns the final transcript text.
+var _ = Describe("emitTranscription", func() {
+	It("streams transcription deltas then a completed event when streaming is enabled", func() {
+		on := true
+		session := &Session{
+			InputAudioTranscription: &types.AudioTranscription{},
+			ModelConfig: &config.ModelConfig{
+				Pipeline: config.Pipeline{Streaming: config.PipelineStreaming{Transcription: &on}},
+			},
+			ModelInterface: &fakeModel{
+				transcribeDeltas: []string{"Hel", "lo", " world"},
+				transcribeFinal:  &schema.TranscriptionResult{Text: "Hello world"},
+			},
+		}
+		t := &fakeTransport{}
+
+		transcript, err := emitTranscription(context.Background(), t, session, "item1", "/tmp/x.wav")
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(transcript).To(Equal("Hello world"))
+		Expect(t.countEvents(types.ServerEventTypeConversationItemInputAudioTranscriptionDelta)).To(Equal(3))
+		Expect(t.countEvents(types.ServerEventTypeConversationItemInputAudioTranscriptionCompleted)).To(Equal(1))
+	})
+
+	It("emits a single completed event with no deltas in unary mode", func() {
+		session := &Session{
+			InputAudioTranscription: &types.AudioTranscription{},
+			ModelConfig:             &config.ModelConfig{}, // streaming off
+			ModelInterface:          &fakeModel{transcribeFinal: &schema.TranscriptionResult{Text: "Hi"}},
+		}
+		t := &fakeTransport{}
+
+		transcript, err := emitTranscription(context.Background(), t, session, "item1", "/tmp/x.wav")
+
+		Expect(err).ToNot(HaveOccurred())
+		Expect(transcript).To(Equal("Hi"))
+		Expect(t.countEvents(types.ServerEventTypeConversationItemInputAudioTranscriptionDelta)).To(Equal(0))
+		Expect(t.countEvents(types.ServerEventTypeConversationItemInputAudioTranscriptionCompleted)).To(Equal(1))
+	})
+})

core/http/endpoints/openai/realtime_webrtc.go

@@ -48,7 +48,8 @@ func RealtimeCalls(application *application.Application) echo.HandlerFunc {
 			return c.JSON(http.StatusInternalServerError, map[string]string{"error": "codec registration failed"})
 		}
 
-		api := webrtc.NewAPI(webrtc.WithMediaEngine(m))
+		se := webRTCSettingEngine(application.ApplicationConfig())
+		api := webrtc.NewAPI(webrtc.WithMediaEngine(m), webrtc.WithSettingEngine(se))
 
 		pc, err := api.NewPeerConnection(webrtc.Configuration{})
 		if err != nil {

core/http/endpoints/openai/realtime_webrtc_ice.go

@@ -0,0 +1,47 @@
+package openai
+
+import (
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/xlog"
+	"github.com/pion/webrtc/v4"
+)
+
+// webRTCSettingEngine builds the pion SettingEngine for /v1/realtime WebRTC.
+//
+// With a default (empty) SettingEngine, pion gathers a host ICE candidate for
+// every local interface. Under Docker host networking that includes bridge
+// addresses (docker0/veth, 172.x) that a remote browser cannot route to; the
+// connection often establishes on a good pair and then drops once ICE consent
+// checks fail on the unreachable ones. The two opt-in knobs below let an
+// operator advertise only the reachable address.
+func webRTCSettingEngine(cfg *config.ApplicationConfig) webrtc.SettingEngine {
+	s := webrtc.SettingEngine{}
+	if cfg == nil {
+		return s
+	}
+	if len(cfg.WebRTCNAT1To1IPs) > 0 {
+		s.SetNAT1To1IPs(cfg.WebRTCNAT1To1IPs, webrtc.ICECandidateTypeHost)
+		xlog.Debug("realtime webrtc: advertising NAT 1:1 host IPs", "ips", cfg.WebRTCNAT1To1IPs)
+	}
+	if filter := iceInterfaceFilter(cfg.WebRTCICEInterfaces); filter != nil {
+		s.SetInterfaceFilter(filter)
+		xlog.Debug("realtime webrtc: restricting ICE interfaces", "interfaces", cfg.WebRTCICEInterfaces)
+	}
+	return s
+}
+
+// iceInterfaceFilter returns an interface allow-list predicate for pion, or nil
+// when no interfaces are configured (pion's default: gather from all).
+func iceInterfaceFilter(allowed []string) func(string) bool {
+	if len(allowed) == 0 {
+		return nil
+	}
+	set := make(map[string]struct{}, len(allowed))
+	for _, name := range allowed {
+		set[name] = struct{}{}
+	}
+	return func(iface string) bool {
+		_, ok := set[iface]
+		return ok
+	}
+}

core/http/endpoints/openai/realtime_webrtc_ice_test.go

@@ -0,0 +1,39 @@
+package openai
+
+import (
+	"github.com/mudler/LocalAI/core/config"
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("webRTC ICE settings", func() {
+	Describe("iceInterfaceFilter", func() {
+		It("returns nil when no interfaces are configured", func() {
+			Expect(iceInterfaceFilter(nil)).To(BeNil())
+			Expect(iceInterfaceFilter([]string{})).To(BeNil())
+		})
+
+		It("admits only the configured interfaces", func() {
+			f := iceInterfaceFilter([]string{"eth0", "wlan0"})
+			Expect(f).NotTo(BeNil())
+			Expect(f("eth0")).To(BeTrue())
+			Expect(f("wlan0")).To(BeTrue())
+			Expect(f("docker0")).To(BeFalse())
+			Expect(f("veth123")).To(BeFalse())
+		})
+	})
+
+	Describe("webRTCSettingEngine", func() {
+		It("does not panic on a nil config", func() {
+			Expect(func() { webRTCSettingEngine(nil) }).NotTo(Panic())
+		})
+
+		It("builds an engine with NAT 1:1 IPs and an interface filter configured", func() {
+			cfg := &config.ApplicationConfig{
+				WebRTCNAT1To1IPs:    []string{"192.168.1.10"},
+				WebRTCICEInterfaces: []string{"eth0"},
+			}
+			Expect(func() { webRTCSettingEngine(cfg) }).NotTo(Panic())
+		})
+	})
+})

core/http/endpoints/openresponses/responses.go

@@ -1356,7 +1356,7 @@ func handleOpenResponsesNonStream(c echo.Context, responseID string, createdAt i
 	thinkingStartToken := reason.DetectThinkingStartToken(template, &cfg.ReasoningConfig)
 
 	// Extract reasoning from result before cleaning
-	reasoningContent, cleanedResult := reason.ExtractReasoningWithConfig(result, thinkingStartToken, cfg.ReasoningConfig)
+	reasoningContent, cleanedResult := reason.ExtractReasoningComplete(result, thinkingStartToken, cfg.ReasoningConfig)
 
 	// Parse tool calls if using functions
 	var outputItems []schema.ORItemField
@@ -1996,7 +1996,7 @@ func handleOpenResponsesStream(c echo.Context, responseID string, createdAt int6
 				finalCleanedResult = extractor.CleanedContent()
 			}
 			if finalReasoning == "" && finalCleanedResult == "" {
-				finalReasoning, finalCleanedResult = reason.ExtractReasoningWithConfig(result, thinkingStartToken, cfg.ReasoningConfig)
+				finalReasoning, finalCleanedResult = reason.ExtractReasoningComplete(result, thinkingStartToken, cfg.ReasoningConfig)
 			}
 
 			// Close reasoning item if it exists and wasn't closed yet
@@ -2493,7 +2493,7 @@ func handleOpenResponsesStream(c echo.Context, responseID string, createdAt int6
 		finalCleanedResult = extractor.CleanedContent()
 	}
 	if finalReasoning == "" && finalCleanedResult == "" {
-		finalReasoning, finalCleanedResult = reason.ExtractReasoningWithConfig(result, thinkingStartToken, cfg.ReasoningConfig)
+		finalReasoning, finalCleanedResult = reason.ExtractReasoningComplete(result, thinkingStartToken, cfg.ReasoningConfig)
 	}
 
 	// Close reasoning item if it exists and wasn't closed yet

core/http/react-ui/src/hooks/useChat.js

@@ -216,6 +216,12 @@ export function useChat(initialModel = '') {
             audio_url: { url: `data:${file.type};base64,${file.base64}` },
           })
           userFiles.push({ name: file.name, type: 'audio' })
+        } else if (file.type?.startsWith('video/')) {
+          messageContent.push({
+            type: 'video_url',
+            video_url: { url: `data:${file.type};base64,${file.base64}` },
+          })
+          userFiles.push({ name: file.name, type: 'video' })
         } else {
 			// Text/PDF files - append to content
 			if (file.textContent) {

core/http/react-ui/src/pages/Backends.jsx

@@ -506,7 +506,10 @@ export default function Backends() {
             <tbody>
               {backends.map((b, idx) => {
                 const op = getBackendOp(b)
-                const isProcessing = !!op
+                // A failed op is intentionally kept in the operations list so the
+                // OperationsBar can surface the error + Dismiss; it must NOT render
+                // as a perpetual "Installing..." spinner here (mirrors Models.jsx).
+                const isProcessing = !!op && !op.error
                 const isExpanded = expandedRow === idx
 
                 return (

core/http/react-ui/src/pages/Chat.jsx

@@ -265,7 +265,7 @@ function UserMessageContent({ content, files }) {
         <div className="chat-message-files">
           {files.map((f, i) => (
             <span key={i} className="chat-file-inline">
-              <i className={`fas ${f.type === 'image' ? 'fa-image' : f.type === 'audio' ? 'fa-headphones' : 'fa-file'}`} />
+              <i className={`fas ${f.type === 'image' ? 'fa-image' : f.type === 'audio' ? 'fa-headphones' : f.type === 'video' ? 'fa-film' : 'fa-file'}`} />
               {f.name}
             </span>
           ))}
@@ -274,6 +274,9 @@ function UserMessageContent({ content, files }) {
       {Array.isArray(content) && content.filter(c => c.type === 'image_url').map((img, i) => (
         <img key={i} src={img.image_url.url} alt="attached" className="chat-inline-image" />
       ))}
+      {Array.isArray(content) && content.filter(c => c.type === 'video_url').map((vid, i) => (
+        <video key={i} src={vid.video_url.url} controls className="chat-inline-video" />
+      ))}
     </>
   )
 }
@@ -711,7 +714,7 @@ export default function Chat() {
     for (const file of e.target.files) {
       const base64 = await fileToBase64(file)
       const entry = { name: file.name, type: file.type, base64 }
-      if (!file.type.startsWith('image/') && !file.type.startsWith('audio/')) {
+      if (!file.type.startsWith('image/') && !file.type.startsWith('audio/') && !file.type.startsWith('video/')) {
         entry.textContent = await file.text().catch(() => '')
       }
       newFiles.push(entry)
@@ -1244,7 +1247,7 @@ export default function Chat() {
           <div className="chat-files">
             {files.map((f, i) => (
               <span key={i} className="chat-file-badge">
-                <i className={`fas ${f.type?.startsWith('image/') ? 'fa-image' : f.type?.startsWith('audio/') ? 'fa-headphones' : 'fa-file'}`} />
+                <i className={`fas ${f.type?.startsWith('image/') ? 'fa-image' : f.type?.startsWith('audio/') ? 'fa-headphones' : f.type?.startsWith('video/') ? 'fa-film' : 'fa-file'}`} />
                 {f.name}
                 <button onClick={() => setFiles(prev => prev.filter((_, idx) => idx !== i))}>
                   <i className="fas fa-xmark" />
@@ -1343,7 +1346,7 @@ export default function Chat() {
               ref={fileInputRef}
               type="file"
               multiple
-              accept="image/*,audio/*,application/pdf,.txt,.md,.csv,.json"
+              accept="image/*,audio/*,video/*,application/pdf,.txt,.md,.csv,.json"
               style={{ display: 'none' }}
               onChange={handleFileChange}
             />

core/http/react-ui/src/pages/Manage.jsx

@@ -12,6 +12,7 @@ import ActionMenu from '../components/ActionMenu'
 import ResourceRow, { ChevronCell, IconCell, StopPropagationCell } from '../components/ResourceRow'
 import { useModels } from '../hooks/useModels'
 import { useGalleryEnrichment } from '../hooks/useGalleryEnrichment'
+import { useOperations } from '../hooks/useOperations'
 import { backendControlApi, modelsApi, backendsApi, systemApi, nodesApi } from '../utils/api'
 import { renderMarkdown } from '../utils/markdown'
 import { safeHref } from '../utils/url'
@@ -126,6 +127,7 @@ export default function Manage() {
   const [activeTab, setActiveTab] = useState(TABS.some(tab => tab.key === initialTab) ? initialTab : 'models')
   const { models, loading: modelsLoading, refetch: refetchModels } = useModels()
   const { enrichModel, enrichBackend } = useGalleryEnrichment()
+  const { operations } = useOperations()
   const [loadedModelIds, setLoadedModelIds] = useState(new Set())
   const [backends, setBackends] = useState([])
   const [backendsLoading, setBackendsLoading] = useState(true)
@@ -258,14 +260,19 @@ export default function Manage() {
     return `${m}m ago`
   })()
 
-  // Fetch available backend upgrades
+  // Refresh installed backends + available upgrades when the Backends tab opens
+  // AND whenever a backend operation settles (operations.length changes as a
+  // reinstall/upgrade completes and drops off the list). Without the op-settle
+  // refresh the installed-version cell and the "update available" badge stay
+  // stale after an upgrade until the user switches tabs - the op looks like it
+  // "did nothing". Mirrors the operations.length watch Backends.jsx uses.
   useEffect(() => {
-    if (activeTab === 'backends') {
-      backendsApi.checkUpgrades()
-        .then(data => setUpgrades(data || {}))
-        .catch(() => {})
-    }
-  }, [activeTab])
+    if (activeTab !== 'backends') return
+    fetchBackends()
+    backendsApi.checkUpgrades()
+      .then(data => setUpgrades(data || {}))
+      .catch(() => {})
+  }, [operations.length, activeTab, fetchBackends])
 
   const handleStopModel = (modelName) => {
     setConfirmDialog({

core/http/react-ui/src/pages/Talk.jsx

@@ -17,6 +17,24 @@ const STATUS_STYLES = {
   error:        { icon: 'fa-solid fa-circle', color: 'var(--color-error)', bg: 'var(--color-error-light)' },
 }
 
+// upsertAssistant merges a streamed transcript fragment into the assistant entry
+// identified by the server's item_id, or appends a new entry if none exists yet.
+// Keying by item_id (not a mutable index tracked across handler/updater
+// boundaries) makes streamed deltas idempotent and order-independent, so React's
+// batching of non-React data-channel events cannot produce a duplicate bubble.
+// mode 'append' adds to the running text; 'replace' sets the final transcript.
+function upsertAssistant(prev, itemId, text, mode) {
+  // Only assistant entries carry an id, and the streaming entry is almost
+  // always the newest — search from the tail so per-delta cost stays constant.
+  const i = prev.findLastIndex(e => e.id === itemId)
+  if (i === -1) {
+    return [...prev, { role: 'assistant', id: itemId, text }]
+  }
+  const next = [...prev]
+  next[i] = { ...next[i], text: mode === 'append' ? next[i].text + text : text }
+  return next
+}
+
 export default function Talk() {
   const { addToast } = useOutletContext()
   const navigate = useNavigate()
@@ -34,7 +52,10 @@ export default function Talk() {
 
   // Transcript
   const [transcript, setTranscript] = useState([])
-  const streamingRef = useRef(null) // tracks the index of the in-progress assistant message
+  // item_id of the assistant message currently streaming — used only to remove
+  // its partial bubble when a response is cancelled (barge-in). The transcript
+  // itself is keyed by item_id via upsertAssistant, not by this ref.
+  const inProgressIdRef = useRef(null)
 
   // Session settings
   const [instructions, setInstructions] = useState(
@@ -227,39 +248,21 @@ export default function Talk() {
         break
       case 'conversation.item.input_audio_transcription.completed':
         if (event.transcript) {
-          streamingRef.current = null
           setTranscript(prev => [...prev, { role: 'user', text: event.transcript }])
         }
         updateStatus('thinking', 'Generating response...')
         break
       case 'response.output_audio_transcript.delta':
         if (event.delta) {
-          setTranscript(prev => {
-            if (streamingRef.current !== null) {
-              const updated = [...prev]
-              updated[streamingRef.current] = {
-                ...updated[streamingRef.current],
-                text: updated[streamingRef.current].text + event.delta,
-              }
-              return updated
-            }
-            streamingRef.current = prev.length
-            return [...prev, { role: 'assistant', text: event.delta }]
-          })
+          inProgressIdRef.current = event.item_id
+          setTranscript(prev => upsertAssistant(prev, event.item_id, event.delta, 'append'))
         }
         break
       case 'response.output_audio_transcript.done':
         if (event.transcript) {
-          setTranscript(prev => {
-            if (streamingRef.current !== null) {
-              const updated = [...prev]
-              updated[streamingRef.current] = { ...updated[streamingRef.current], text: event.transcript }
-              return updated
-            }
-            return [...prev, { role: 'assistant', text: event.transcript }]
-          })
+          setTranscript(prev => upsertAssistant(prev, event.item_id, event.transcript, 'replace'))
         }
-        streamingRef.current = null
+        inProgressIdRef.current = null
         break
       case 'response.output_audio.delta':
         updateStatus('speaking', 'Speaking...')
@@ -281,7 +284,7 @@ export default function Talk() {
           // Pretty-print JSON for readability; fall back to raw string.
           try { preview = JSON.stringify(JSON.parse(preview), null, 2) } catch (_) { /* keep raw */ }
           setTranscript(prev => [...prev, { role: 'tool_result', text: preview }])
-          streamingRef.current = null  // tool result ends the current assistant text run
+          inProgressIdRef.current = null // tool result ends the current assistant text run
         }
         break
       }
@@ -290,9 +293,20 @@ export default function Talk() {
         // conversation.item.create + response.create when it's done.
         handleFunctionCall(event)
         break
-      case 'response.done':
+      case 'response.done': {
+        // A cancelled response (barge-in / interruption) leaves a partial,
+        // incrementally-streamed assistant bubble behind. The server discards
+        // the interrupted item from history; mirror that here (remove the
+        // in-progress assistant entry by item_id) so the regenerated reply
+        // doesn't show up as a second assistant message.
+        if (event.response?.status === 'cancelled' && inProgressIdRef.current) {
+          const id = inProgressIdRef.current
+          inProgressIdRef.current = null
+          setTranscript(prev => prev.filter(e => e.id !== id))
+        }
         updateStatus('listening', 'Listening...')
         break
+      }
       case 'error':
         hasErrorRef.current = true
         updateStatus('error', 'Error: ' + (event.error?.message || 'Unknown error'))
@@ -789,7 +803,7 @@ export default function Talk() {
               const iconColor = isToolCall || isToolResult ? 'var(--color-text-secondary)'
                               : isUser ? 'var(--color-primary)' : 'var(--color-accent)'
               return (
-                <div key={i} style={{ display: 'flex', alignItems: 'flex-start', gap: 'var(--spacing-xs)' }}>
+                <div key={entry.id || i} style={{ display: 'flex', alignItems: 'flex-start', gap: 'var(--spacing-xs)' }}>
                   <i className={iconClass} style={{ color: iconColor, marginTop: 3, flexShrink: 0, fontSize: '0.75rem' }} />
                   <p style={{
                     margin: 0,

core/services/agentpool/agent_pool.go

@@ -466,10 +466,11 @@ func (s *AgentPoolService) Chat(name, message string) (string, error) {
 				s.collectAndCopyMetadata(metadata, chatUserID)
 			}
 
+			content := s.appendLocalAGIKBCitations(response.Response, name, message, response.State)
 			msg := map[string]any{
 				"id":        messageID + "-agent",
 				"sender":    "agent",
-				"content":   response.Response,
+				"content":   content,
 				"timestamp": time.Now().Format(time.RFC3339),
 			}
 			if len(metadata) > 0 {
@@ -489,6 +490,79 @@ func (s *AgentPoolService) Chat(name, message string) (string, error) {
 	return messageID, nil
 }
 
+func (s *AgentPoolService) appendLocalAGIKBCitations(response, agentKey, message string, states []coreTypes.ActionState) string {
+	if strings.TrimSpace(response) == "" {
+		return response
+	}
+
+	userID, collection := splitAgentKey(agentKey)
+	cfg := s.localAGI.pool.GetConfig(agentKey)
+	if cfg == nil || !cfg.EnableKnowledgeBase {
+		return response
+	}
+
+	citations := kbCitationsFromActionStates(states)
+	if len(citations) == 0 && cfg.KBAutoSearch {
+		maxResults := cfg.KnowledgeBaseResults
+		if maxResults <= 0 {
+			maxResults = 5
+		}
+		ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+		defer cancel()
+		kbResult := agents.KBAutoSearchPrompt(ctx, s.apiURL, s.apiKey, collection, message, maxResults, userID)
+		citations = kbResult.Citations
+	}
+
+	return agents.AppendKBCitations(response, collection, userID, citations)
+}
+
+func splitAgentKey(agentKey string) (userID, name string) {
+	if uid, n, ok := strings.Cut(agentKey, ":"); ok {
+		return uid, n
+	}
+	return "", agentKey
+}
+
+func kbCitationsFromActionStates(states []coreTypes.ActionState) []agents.KBCitation {
+	var citations []agents.KBCitation
+	for _, state := range states {
+		citations = append(citations, kbCitationsFromMetadata(state.Metadata)...)
+	}
+	return citations
+}
+
+func kbCitationsFromMetadata(metadata map[string]any) []agents.KBCitation {
+	if len(metadata) == 0 {
+		return nil
+	}
+
+	fileName := metadata["file_name"]
+	source := metadata["source"]
+	if fileName == nil && source == nil {
+		return nil
+	}
+
+	citation := agents.KBCitation{
+		FileName: metadataString(fileName),
+		EntryKey: metadataString(source),
+	}
+	if citation.FileName == "" && citation.EntryKey == "" {
+		return nil
+	}
+	return []agents.KBCitation{citation}
+}
+
+func metadataString(value any) string {
+	switch v := value.(type) {
+	case string:
+		return v
+	case fmt.Stringer:
+		return v.String()
+	default:
+		return ""
+	}
+}
+
 // userOutputsDir returns the per-user outputs directory, creating it if needed.
 // If userID is empty, falls back to the shared outputs directory.
 func (s *AgentPoolService) userOutputsDir(userID string) string {

core/services/agents/citations.go

@@ -0,0 +1,127 @@
+package agents
+
+import (
+	"fmt"
+	"net/url"
+	"strings"
+	"sync"
+)
+
+type kbCitationList struct {
+	mu        sync.Mutex
+	citations []KBCitation
+}
+
+func (l *kbCitationList) AddKBCitations(citations []KBCitation) {
+	if len(citations) == 0 {
+		return
+	}
+	l.mu.Lock()
+	defer l.mu.Unlock()
+	l.citations = append(l.citations, citations...)
+}
+
+func (l *kbCitationList) Citations() []KBCitation {
+	l.mu.Lock()
+	defer l.mu.Unlock()
+	out := make([]KBCitation, len(l.citations))
+	copy(out, l.citations)
+	return out
+}
+
+// AppendKBCitations appends a markdown Sources block for KB citations.
+func AppendKBCitations(response, collection, userID string, citations []KBCitation) string {
+	if strings.TrimSpace(response) == "" || len(citations) == 0 {
+		return response
+	}
+
+	var lines []string
+	seen := make(map[string]struct{})
+	for _, citation := range citations {
+		key := strings.TrimSpace(citation.EntryKey)
+		if key == "" {
+			key = strings.TrimSpace(citation.FileName)
+		}
+		if key == "" {
+			continue
+		}
+		if _, ok := seen[key]; ok {
+			continue
+		}
+		seen[key] = struct{}{}
+
+		displayName := kbCitationDisplayName(citation)
+		if displayName == "" {
+			continue
+		}
+
+		sourceURL := kbCitationRawFileURL(collection, citation.EntryKey, userID)
+		number := len(lines) + 1
+		if sourceURL == "" {
+			lines = append(lines, fmt.Sprintf("[%d] %s", number, displayName))
+			continue
+		}
+		lines = append(lines, fmt.Sprintf("[%d] [%s](%s)", number, escapeMarkdownLinkText(displayName), sourceURL))
+	}
+
+	if len(lines) == 0 {
+		return response
+	}
+
+	var sb strings.Builder
+	sb.WriteString(strings.TrimRight(response, "\n"))
+	sb.WriteString("\n\nSources:\n")
+	for _, line := range lines {
+		sb.WriteString(line)
+		sb.WriteString("\n")
+	}
+	return strings.TrimRight(sb.String(), "\n")
+}
+
+func kbCitationDisplayName(citation KBCitation) string {
+	if fileName := strings.TrimSpace(citation.FileName); fileName != "" {
+		return fileName
+	}
+
+	segments := strings.Split(strings.Trim(strings.TrimSpace(citation.EntryKey), "/"), "/")
+	for i := len(segments) - 1; i >= 0; i-- {
+		if segment := strings.TrimSpace(segments[i]); segment != "" {
+			return segment
+		}
+	}
+	return ""
+}
+
+func kbCitationRawFileURL(collection, entryKey, userID string) string {
+	collection = strings.TrimSpace(collection)
+	entryKey = strings.Trim(strings.TrimSpace(entryKey), "/")
+	if collection == "" || entryKey == "" {
+		return ""
+	}
+
+	var escapedEntrySegments []string
+	for _, segment := range strings.Split(entryKey, "/") {
+		if segment == "" {
+			continue
+		}
+		escapedEntrySegments = append(escapedEntrySegments, url.PathEscape(segment))
+	}
+	if len(escapedEntrySegments) == 0 {
+		return ""
+	}
+
+	sourceURL := "/api/agents/collections/" + url.PathEscape(collection) + "/entries-raw/" + strings.Join(escapedEntrySegments, "/")
+	if userID != "" {
+		query := url.Values{}
+		query.Set("user_id", userID)
+		sourceURL += "?" + query.Encode()
+	}
+	return sourceURL
+}
+
+func escapeMarkdownLinkText(text string) string {
+	text = strings.ReplaceAll(text, `\`, `\\`)
+	text = strings.ReplaceAll(text, "[", `\[`)
+	text = strings.ReplaceAll(text, "]", `\]`)
+	return text
+}

core/services/agents/executor.go

@@ -167,10 +167,12 @@ func ExecuteChatWithLLM(ctx context.Context, llm cogito.LLM, cfg *AgentConfig, m
 		}
 	}
 
+	kbCitations := &kbCitationList{}
 	if cfg.EnableKnowledgeBase && (kbMode == KBModeAutoSearch || kbMode == KBModeBoth) {
-		kbResults := KBAutoSearchPrompt(ctx, effectiveURL, effectiveKey, cfg.Name, message, cfg.KnowledgeBaseResults, userID)
-		if kbResults != "" {
-			fragment = fragment.AddMessage(cogito.SystemMessageRole, kbResults)
+		kbResult := KBAutoSearchPrompt(ctx, effectiveURL, effectiveKey, cfg.Name, message, cfg.KnowledgeBaseResults, userID)
+		if kbResult.Prompt != "" {
+			fragment = fragment.AddMessage(cogito.SystemMessageRole, kbResult.Prompt)
+			kbCitations.AddKBCitations(kbResult.Citations)
 		}
 	}
 
@@ -197,7 +199,7 @@ func ExecuteChatWithLLM(ctx context.Context, llm cogito.LLM, cfg *AgentConfig, m
 		}
 		cogitoOpts = append(cogitoOpts, cogito.WithTools(
 			cogito.NewToolDefinition(
-				KBSearchMemoryTool{APIURL: effectiveURL, APIKey: effectiveKey, Collection: cfg.Name, MaxResults: kbResults, UserID: userID},
+				KBSearchMemoryTool{APIURL: effectiveURL, APIKey: effectiveKey, Collection: cfg.Name, MaxResults: kbResults, UserID: userID, CitationCollector: kbCitations},
 				KBSearchMemoryArgs{},
 				"search_memory",
 				"Search the knowledge base for relevant information",
@@ -336,6 +338,8 @@ func ExecuteChatWithLLM(ctx context.Context, llm cogito.LLM, cfg *AgentConfig, m
 	if cfg.StripThinkingTags && response != "" {
 		response = stripThinkingTags(response)
 	}
+	responseForMemory := response
+	response = AppendKBCitations(response, cfg.Name, userID, kbCitations.Citations())
 
 	// Save conversation to KB when long-term memory is enabled.
 	// Use a detached context: the parent ctx may be cancelled (e.g. in distributed
@@ -344,7 +348,7 @@ func ExecuteChatWithLLM(ctx context.Context, llm cogito.LLM, cfg *AgentConfig, m
 		go func() {
 			ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 			defer cancel()
-			saveConversationToKB(ctx, llm, effectiveURL, effectiveKey, cfg, message, response, userID)
+			saveConversationToKB(ctx, llm, effectiveURL, effectiveKey, cfg, message, responseForMemory, userID)
 		}()
 	}
 

core/services/agents/executor_test.go

@@ -2,6 +2,8 @@ package agents
 
 import (
 	"context"
+	"net/http"
+	"net/http/httptest"
 	"sync"
 	"sync/atomic"
 
@@ -36,6 +38,34 @@ func (m *mockLLM) CreateChatCompletion(ctx context.Context, req openai.ChatCompl
 	}, cogito.LLMUsage{}, nil
 }
 
+type toolCallingMockLLM struct {
+	createResponses []openai.ChatCompletionResponse
+	askResponse     string
+	callCount       atomic.Int32
+}
+
+func (m *toolCallingMockLLM) Ask(ctx context.Context, f cogito.Fragment) (cogito.Fragment, error) {
+	m.callCount.Add(1)
+	return f.AddMessage(cogito.AssistantMessageRole, m.askResponse), nil
+}
+
+func (m *toolCallingMockLLM) CreateChatCompletion(ctx context.Context, req openai.ChatCompletionRequest) (cogito.LLMReply, cogito.LLMUsage, error) {
+	idx := int(m.callCount.Add(1)) - 1
+	if idx >= len(m.createResponses) {
+		return cogito.LLMReply{
+			ChatCompletionResponse: openai.ChatCompletionResponse{
+				Choices: []openai.ChatCompletionChoice{{
+					Message: openai.ChatCompletionMessage{
+						Role:    "assistant",
+						Content: "No more tools needed.",
+					},
+				}},
+			},
+		}, cogito.LLMUsage{}, nil
+	}
+	return cogito.LLMReply{ChatCompletionResponse: m.createResponses[idx]}, cogito.LLMUsage{}, nil
+}
+
 // statusCollector records status callbacks in a thread-safe way.
 type statusCollector struct {
 	mu       sync.Mutex
@@ -73,6 +103,74 @@ var _ = DescribeTable("stripThinkingTags",
 	Entry("adjacent tag pairs", "<thinking>a</thinking><thinking>b</thinking>", ""),
 )
 
+var _ = DescribeTable("appendKBCitations",
+	func(response, collection, userID string, citations []KBCitation, want string) {
+		Expect(AppendKBCitations(response, collection, userID, citations)).To(Equal(want))
+	},
+	Entry("leaves responses without citations unchanged",
+		"answer",
+		"agent",
+		"",
+		nil,
+		"answer",
+	),
+	Entry("leaves blank responses unchanged",
+		"",
+		"agent",
+		"",
+		[]KBCitation{{FileName: "source.pdf", EntryKey: "uuid/source.pdf"}},
+		"",
+	),
+	Entry("appends clickable source links",
+		"answer",
+		"my-agent",
+		"",
+		[]KBCitation{{FileName: "new feature.pdf", EntryKey: "uuid/new feature.pdf"}},
+		"answer\n\nSources:\n[1] [new feature.pdf](/api/agents/collections/my-agent/entries-raw/uuid/new%20feature.pdf)",
+	),
+	Entry("deduplicates citations by entry key",
+		"answer",
+		"agent",
+		"",
+		[]KBCitation{
+			{FileName: "first.pdf", EntryKey: "uuid/shared.pdf"},
+			{FileName: "second.pdf", EntryKey: "uuid/shared.pdf"},
+		},
+		"answer\n\nSources:\n[1] [first.pdf](/api/agents/collections/agent/entries-raw/uuid/shared.pdf)",
+	),
+	Entry("uses plain text when entry key is missing",
+		"answer",
+		"agent",
+		"",
+		[]KBCitation{{FileName: "source.pdf"}},
+		"answer\n\nSources:\n[1] source.pdf",
+	),
+	Entry("uses entry basename when filename is missing",
+		"answer",
+		"agent",
+		"",
+		[]KBCitation{{EntryKey: "uuid/source.pdf"}},
+		"answer\n\nSources:\n[1] [source.pdf](/api/agents/collections/agent/entries-raw/uuid/source.pdf)",
+	),
+	Entry("adds user id query when present",
+		"answer",
+		"agent",
+		"user 1",
+		[]KBCitation{{FileName: "source.pdf", EntryKey: "uuid/source.pdf"}},
+		"answer\n\nSources:\n[1] [source.pdf](/api/agents/collections/agent/entries-raw/uuid/source.pdf?user_id=user+1)",
+	),
+	Entry("escapes collection, path segments, and markdown link text",
+		"answer",
+		"agent one",
+		"",
+		[]KBCitation{{FileName: "source [draft].pdf", EntryKey: "uuid/source [draft].pdf"}},
+		`answer
+
+Sources:
+[1] [source \[draft\].pdf](/api/agents/collections/agent%20one/entries-raw/uuid/source%20%5Bdraft%5D.pdf)`,
+	),
+)
+
 var _ = Describe("ExecuteChatWithLLM", func() {
 	var (
 		ctx context.Context
@@ -184,6 +282,150 @@ var _ = Describe("ExecuteChatWithLLM", func() {
 		})
 	})
 
+	Context("knowledge base citations", func() {
+		It("appends KB sources to the returned response and callback message", func() {
+			mux := http.NewServeMux()
+			mux.HandleFunc("/api/agents/collections/kb-agent/search", func(w http.ResponseWriter, r *http.Request) {
+				Expect(r.URL.Query().Get("user_id")).To(Equal("user-1"))
+				w.Header().Set("Content-Type", "application/json")
+				_, _ = w.Write([]byte(`{
+					"results": [
+						{
+							"content": "KB content",
+							"id": "result-1",
+							"similarity": 0.99,
+							"metadata": {
+								"file_name": "new feature.pdf",
+								"source": "uuid/new feature.pdf"
+							}
+						}
+					],
+					"count": 1
+				}`))
+			})
+			server := httptest.NewServer(mux)
+			defer server.Close()
+
+			var msgContent string
+			cb.OnMessage = func(sender, content, messageID string) {
+				msgContent = content
+			}
+
+			llm := &mockLLM{response: "agent reply"}
+			cfg := &AgentConfig{
+				Name:                "kb-agent",
+				Model:               "test-model",
+				EnableKnowledgeBase: true,
+				KBMode:              KBModeAutoSearch,
+			}
+
+			result, err := ExecuteChatWithLLM(ctx, llm, cfg, "hello", cb, ExecuteChatOpts{
+				APIURL: server.URL,
+				UserID: "user-1",
+			})
+			Expect(err).ToNot(HaveOccurred())
+			Expect(result).To(Equal("agent reply\n\nSources:\n[1] [new feature.pdf](/api/agents/collections/kb-agent/entries-raw/uuid/new%20feature.pdf?user_id=user-1)"))
+			Expect(msgContent).To(Equal(result))
+		})
+
+		It("collects citations from the search_memory tool", func() {
+			mux := http.NewServeMux()
+			mux.HandleFunc("/api/agents/collections/kb-agent/search", func(w http.ResponseWriter, r *http.Request) {
+				w.Header().Set("Content-Type", "application/json")
+				_, _ = w.Write([]byte(`{
+					"results": [
+						{
+							"content": "Tool KB content",
+							"id": "result-1",
+							"similarity": 0.99,
+							"metadata": {
+								"file_name": "tool source.pdf",
+								"source": "uuid/tool source.pdf"
+							}
+						}
+					],
+					"count": 1
+				}`))
+			})
+			server := httptest.NewServer(mux)
+			defer server.Close()
+
+			collector := &kbCitationList{}
+			tool := KBSearchMemoryTool{
+				APIURL:            server.URL,
+				Collection:        "kb-agent",
+				CitationCollector: collector,
+			}
+
+			result, _, err := tool.Run(KBSearchMemoryArgs{Query: "hello"})
+			Expect(err).ToNot(HaveOccurred())
+			Expect(result).To(ContainSubstring("Tool KB content"))
+			Expect(collector.Citations()).To(Equal([]KBCitation{{FileName: "tool source.pdf", EntryKey: "uuid/tool source.pdf"}}))
+		})
+
+		It("appends KB sources found through tools-only search_memory calls", func() {
+			mux := http.NewServeMux()
+			mux.HandleFunc("/api/agents/collections/kb-agent/search", func(w http.ResponseWriter, r *http.Request) {
+				Expect(r.URL.Query().Get("user_id")).To(Equal("user-1"))
+				w.Header().Set("Content-Type", "application/json")
+				_, _ = w.Write([]byte(`{
+					"results": [
+						{
+							"content": "Tool KB content",
+							"id": "result-1",
+							"similarity": 0.99,
+							"metadata": {
+								"file_name": "tool source.pdf",
+								"source": "uuid/tool source.pdf"
+							}
+						}
+					],
+					"count": 1
+				}`))
+			})
+			server := httptest.NewServer(mux)
+			defer server.Close()
+
+			llm := &toolCallingMockLLM{
+				askResponse: "agent reply from tool context",
+				createResponses: []openai.ChatCompletionResponse{
+					{
+						Choices: []openai.ChatCompletionChoice{
+							{
+								Message: openai.ChatCompletionMessage{
+									Role: "assistant",
+									ToolCalls: []openai.ToolCall{
+										{
+											ID:   "call-1",
+											Type: openai.ToolTypeFunction,
+											Function: openai.FunctionCall{
+												Name:      "search_memory",
+												Arguments: `{"query":"hello"}`,
+											},
+										},
+									},
+								},
+							},
+						},
+					},
+				},
+			}
+			cfg := &AgentConfig{
+				Name:                "kb-agent",
+				Model:               "test-model",
+				EnableKnowledgeBase: true,
+				KBMode:              KBModeTools,
+			}
+
+			result, err := ExecuteChatWithLLM(ctx, llm, cfg, "hello", cb, ExecuteChatOpts{
+				APIURL: server.URL,
+				UserID: "user-1",
+			})
+			Expect(err).ToNot(HaveOccurred())
+			Expect(result).To(Equal("agent reply from tool context\n\nSources:\n[1] [tool source.pdf](/api/agents/collections/kb-agent/entries-raw/uuid/tool%20source.pdf?user_id=user-1)"))
+		})
+	})
+
 	Context("context cancellation", func() {
 		It("returns an error when context is already cancelled", func() {
 			cancelledCtx, cancel := context.WithCancel(ctx)

core/services/agents/knowledge.go

@@ -8,6 +8,7 @@ import (
 	"io"
 	"mime/multipart"
 	"net/http"
+	"net/url"
 	"strings"
 	"time"
 
@@ -17,10 +18,19 @@ import (
 	"github.com/mudler/LocalAI/pkg/httpclient"
 )
 
+// Metadata keys populated by localrecall for every stored chunk. The original
+// upload file name lives under file_name (used for display); source holds the
+// collection entry key ("<uuid>/<filename>") used to build the raw-file URL.
+const (
+	kbMetadataFileName = "file_name"
+	kbMetadataSource   = "source"
+)
+
 // KBSearchResult represents a search result from the knowledge base.
+// Field names mirror the collection search endpoint's JSON response.
 type KBSearchResult struct {
 	Content    string            `json:"content"`
-	Score      float64           `json:"score"`
+	ID         string            `json:"id"`
 	Similarity float64           `json:"similarity"`
 	Metadata   map[string]string `json:"metadata"`
 }
@@ -31,22 +41,48 @@ type kbSearchResponse struct {
 	Count   int              `json:"count"`
 }
 
-// KBAutoSearchPrompt queries the knowledge base with the user's message
-// and returns a system prompt block with relevant results.
+// KBCitation is a single source document that a KB search drew from. Citations
+// travel alongside the prompt as structured data so the consumer (and UI) can
+// render clickable source links, independent of what the model writes inline.
+type KBCitation struct {
+	// FileName is the original uploaded file name, for display (e.g. "report.pdf").
+	FileName string `json:"file_name"`
+	// EntryKey is the collection entry identifier ("<uuid>/<filename>"), used to
+	// build the raw-file URL and as the de-duplication key.
+	EntryKey string `json:"entry_key"`
+}
+
+// KBSearchContext is the result of an auto-search against the knowledge base:
+// the system-prompt block to feed the model, plus the de-duplicated list of
+// source documents the results were drawn from.
+type KBSearchContext struct {
+	Prompt    string       `json:"prompt"`
+	Citations []KBCitation `json:"citations"`
+}
+
+// KBCitationCollector receives source citations found during KB searches.
+type KBCitationCollector interface {
+	AddKBCitations([]KBCitation)
+}
+
+// KBAutoSearchPrompt queries the knowledge base with the user's message and
+// returns a KBSearchContext: a system prompt block with the relevant results
+// plus the de-duplicated source citations those results came from.
 // Uses LocalAI's collection search endpoint via the API.
-func KBAutoSearchPrompt(ctx context.Context, apiURL, apiKey, collection, query string, maxResults int, userID string) string {
+func KBAutoSearchPrompt(ctx context.Context, apiURL, apiKey, collection, query string, maxResults int, userID string) KBSearchContext {
 	if collection == "" || query == "" {
-		return ""
+		return KBSearchContext{}
 	}
 
 	if maxResults <= 0 {
 		maxResults = 5
 	}
 
-	// Call LocalAI's collection search API
-	searchURL := strings.TrimRight(apiURL, "/") + "/api/agents/collections/" + collection + "/search"
+	searchURL := strings.TrimRight(apiURL, "/") + "/api/agents/collections/" + url.PathEscape(collection) + "/search"
 	if userID != "" {
-		searchURL += "?user_id=" + userID
+		query := url.Values{}
+		query.Set("user_id", userID)
+		searchURL += "?" + query.Encode()
 	}
 	reqBody, _ := json.Marshal(map[string]any{
 		"query":       query,
@@ -56,7 +92,7 @@ func KBAutoSearchPrompt(ctx context.Context, apiURL, apiKey, collection, query s
 	req, err := http.NewRequestWithContext(ctx, http.MethodPost, searchURL, strings.NewReader(string(reqBody)))
 	if err != nil {
 		xlog.Warn("KB auto-search: failed to create request", "error", err)
-		return ""
+		return KBSearchContext{}
 	}
 	req.Header.Set("Content-Type", "application/json")
 	if apiKey != "" {
@@ -66,41 +102,70 @@ func KBAutoSearchPrompt(ctx context.Context, apiURL, apiKey, collection, query s
 	resp, err := httpclient.New().Do(req)
 	if err != nil {
 		xlog.Warn("KB auto-search: request failed", "error", err)
-		return ""
+		return KBSearchContext{}
 	}
 	defer resp.Body.Close()
 
 	if resp.StatusCode != http.StatusOK {
 		body, _ := io.ReadAll(resp.Body)
 		xlog.Warn("KB auto-search: non-200 response", "status", resp.StatusCode, "body", string(body))
-		return ""
+		return KBSearchContext{}
 	}
 
 	var searchResp kbSearchResponse
 	if err := json.NewDecoder(resp.Body).Decode(&searchResp); err != nil {
 		xlog.Warn("KB auto-search: failed to decode response", "error", err)
-		return ""
+		return KBSearchContext{}
 	}
 
 	if len(searchResp.Results) == 0 {
-		return ""
+		return KBSearchContext{}
 	}
 
-	// Format results as a system prompt block (same format as LocalAGI)
+	// Build the system prompt block, labelling each chunk with its source file
+	// so the model can attribute inline, and collect the structured citations.
 	var sb strings.Builder
 	sb.WriteString("Given the user input you have the following in memory:\n")
-	for i, r := range searchResp.Results {
-		sb.WriteString(fmt.Sprintf("- %s", r.Content))
-		if len(r.Metadata) > 0 {
-			meta, _ := json.Marshal(r.Metadata)
-			sb.WriteString(fmt.Sprintf(" (%s)", string(meta)))
+
+	var citations []KBCitation
+	seen := make(map[string]struct{})
+
+	for _, r := range searchResp.Results {
+		fileName := r.Metadata[kbMetadataFileName]
+		source := r.Metadata[kbMetadataSource]
+
+		label := fileName
+		if label == "" {
+			label = "unknown"
 		}
-		if i < len(searchResp.Results)-1 {
-			sb.WriteString("\n")
+		sb.WriteString(fmt.Sprintf("[Source: %s]\n%s\n", label, r.Content))
+
+		// Citations are de-duplicated per source document: many chunks from the
+		// same file share one source key, so a file is listed only once. Skip
+		// results with no source key — they cannot be linked back to a document.
+		dedupKey := source
+		if dedupKey == "" {
+			dedupKey = fileName
 		}
+		if dedupKey == "" {
+			continue
+		}
+		if _, ok := seen[dedupKey]; ok {
+			continue
+		}
+		seen[dedupKey] = struct{}{}
+		citations = append(citations, KBCitation{
+			FileName: fileName,
+			EntryKey: source,
+		})
 	}
 
-	return sb.String()
+	sb.WriteString("When answering, cite sources using [Source: filename].")
+
+	return KBSearchContext{
+		Prompt:    sb.String(),
+		Citations: citations,
+	}
 }
 
 // KBSearchMemoryArgs defines the arguments for the search_memory tool.
@@ -110,21 +175,25 @@ type KBSearchMemoryArgs struct {
 
 // KBSearchMemoryTool implements the search_memory MCP tool.
 type KBSearchMemoryTool struct {
-	APIURL     string
-	APIKey     string
-	Collection string
-	MaxResults int
-	UserID     string
+	APIURL            string
+	APIKey            string
+	Collection        string
+	MaxResults        int
+	UserID            string
+	CitationCollector KBCitationCollector
 }
 
 func (t KBSearchMemoryTool) Run(args KBSearchMemoryArgs) (string, any, error) {
 	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
 	defer cancel()
 	result := KBAutoSearchPrompt(ctx, t.APIURL, t.APIKey, t.Collection, args.Query, t.MaxResults, t.UserID)
-	if result == "" {
+	if result.Prompt == "" {
 		return "No results found.", nil, nil
 	}
-	return result, nil, nil
+	if t.CitationCollector != nil {
+		t.CitationCollector.AddKBCitations(result.Citations)
+	}
+	return result.Prompt, nil, nil
 }
 
 // KBAddMemoryArgs defines the arguments for the add_memory tool.
@@ -156,9 +225,11 @@ func (t KBAddMemoryTool) Run(args KBAddMemoryArgs) (string, any, error) {
 
 // KBStoreContent uploads text content to a collection via the multipart upload API.
 func KBStoreContent(ctx context.Context, apiURL, apiKey, collection, content, userID string) error {
-	uploadURL := strings.TrimRight(apiURL, "/") + "/api/agents/collections/" + collection + "/upload"
+	uploadURL := strings.TrimRight(apiURL, "/") + "/api/agents/collections/" + url.PathEscape(collection) + "/upload"
 	if userID != "" {
-		uploadURL += "?user_id=" + userID
+		query := url.Values{}
+		query.Set("user_id", userID)
+		uploadURL += "?" + query.Encode()
 	}
 
 	// Build multipart form with the text content as a file

core/services/distributed/gallery.go

@@ -180,18 +180,21 @@ func (s *GalleryStore) Cancel(id string) error {
 	return s.UpdateStatus(id, "cancelled", "")
 }
 
-// CleanStale marks abandoned in-progress operations as failed.
-// Should be called on startup to recover from crashed instances that
-// left records in pending/downloading/processing state.
-func (s *GalleryStore) CleanStale(age time.Duration) error {
+// CleanStale marks abandoned in-progress operations as failed and returns the
+// number of rows reaped. Called on startup AND periodically to recover from
+// crashed/restarted instances that left records in pending/downloading/
+// processing state — an op orphaned after startup would otherwise linger
+// "processing" until the next restart.
+func (s *GalleryStore) CleanStale(age time.Duration) (int64, error) {
 	cutoff := time.Now().Add(-age)
-	return s.db.Model(&GalleryOperationRecord{}).
+	res := s.db.Model(&GalleryOperationRecord{}).
 		Where("updated_at < ? AND status IN ?", cutoff, activeStatuses).
 		Updates(map[string]any{
 			"status":     "failed",
-			"error":      "stale operation cleaned up on startup",
+			"error":      "stale operation reaped (abandoned by a crashed or restarted instance)",
 			"updated_at": time.Now(),
-		}).Error
+		})
+	return res.RowsAffected, res.Error
 }
 
 // CleanOld removes operations older than the given duration.

core/services/galleryop/backends.go

@@ -71,7 +71,7 @@ func (g *GalleryService) backendHandler(op *ManagementOp[gallery.GalleryBackend,
 
 	var err error
 	if op.Upgrade {
-		err = g.backendManager.UpgradeBackend(ctx, op.GalleryElementName, progressCallback)
+		err = g.backendManager.UpgradeBackend(ctx, op.ID, op.GalleryElementName, progressCallback)
 	} else if op.Delete {
 		err = g.backendManager.DeleteBackend(op.GalleryElementName)
 	} else {

core/services/galleryop/cancel_persist_test.go

@@ -0,0 +1,106 @@
+package galleryop_test
+
+import (
+	"context"
+	"time"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/services/distributed"
+	"github.com/mudler/LocalAI/core/services/galleryop"
+	"github.com/mudler/LocalAI/core/services/testutil"
+)
+
+// Reproduces "a cancelled/orphaned op resurrects as 'processing' after a pod
+// restart". CancelOperation flipped the in-memory status to cancelled and
+// broadcast a NATS event, but never persisted the terminal status to the
+// gallery store. On the next replica restart the still-"pending" row hydrated
+// straight back into processingBackends and the UI spun again. CancelOperation
+// must persist the cancellation so it survives a restart.
+var _ = Describe("GalleryService.CancelOperation persistence", func() {
+	It("persists the cancelled status to the gallery store", func() {
+		db := testutil.SetupTestDB()
+		store, err := distributed.NewGalleryStore(db)
+		Expect(err).ToNot(HaveOccurred())
+
+		// Seed an in-flight op as if a replica was mid-install.
+		Expect(store.Create(&distributed.GalleryOperationRecord{
+			ID:                 "op-cancel",
+			GalleryElementName: "llama-cpp-development",
+			OpType:             "backend_install",
+			Status:             "pending",
+			Progress:           0,
+		})).To(Succeed())
+
+		svc := galleryop.NewGalleryService(&config.ApplicationConfig{}, nil)
+		svc.SetGalleryStore(store)
+		// Make the op locally cancellable so CancelOperation proceeds.
+		svc.StoreCancellation("op-cancel", context.CancelFunc(func() {}))
+
+		Expect(svc.CancelOperation("op-cancel")).To(Succeed())
+
+		// The persisted row must now be terminal — otherwise it re-hydrates as
+		// pending on the next restart.
+		rec, err := store.Get("op-cancel")
+		Expect(err).ToNot(HaveOccurred())
+		Expect(rec.Status).To(Equal("cancelled"))
+
+		// And a fresh service hydrating from the store must NOT see it as active.
+		fresh := galleryop.NewGalleryService(&config.ApplicationConfig{}, nil)
+		fresh.SetGalleryStore(store)
+		Expect(fresh.Hydrate()).To(Succeed())
+		Expect(fresh.GetStatus("op-cancel")).To(BeNil(),
+			"a cancelled op must not hydrate back as active after a restart")
+	})
+})
+
+// Reproduces "an op orphaned by a replica that died mid-flight stays 'pending'
+// forever". CleanStale (which marks abandoned active ops failed) only ran once
+// on startup, so an op orphaned AFTER startup was never reaped until the next
+// restart. The service must reap stale ops on an interval, not just at boot.
+var _ = Describe("GalleryService.ReapStaleOperations", func() {
+	It("marks abandoned active ops terminal once they pass the age cutoff", func() {
+		db := testutil.SetupTestDB()
+		store, err := distributed.NewGalleryStore(db)
+		Expect(err).ToNot(HaveOccurred())
+
+		Expect(store.Create(&distributed.GalleryOperationRecord{
+			ID:                 "orphan-op",
+			GalleryElementName: "llama-cpp-development",
+			OpType:             "backend_install",
+			Status:             "pending",
+			Progress:           0,
+		})).To(Succeed())
+		// Force the row's updated_at into the past so it is older than the cutoff.
+		Expect(db.Exec(
+			"UPDATE gallery_operations SET updated_at = ? WHERE id = ?",
+			time.Now().Add(-1*time.Hour), "orphan-op",
+		).Error).To(Succeed())
+
+		// A fresh, still-progressing op must NOT be reaped.
+		Expect(store.Create(&distributed.GalleryOperationRecord{
+			ID:                 "live-op",
+			GalleryElementName: "vllm-development",
+			OpType:             "backend_install",
+			Status:             "downloading",
+			Progress:           50,
+		})).To(Succeed())
+
+		svc := galleryop.NewGalleryService(&config.ApplicationConfig{}, nil)
+		svc.SetGalleryStore(store)
+
+		reaped, err := svc.ReapStaleOperations(30 * time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(reaped).To(Equal(int64(1)))
+
+		orphan, err := store.Get("orphan-op")
+		Expect(err).ToNot(HaveOccurred())
+		Expect(orphan.Status).To(Equal("failed"))
+
+		live, err := store.Get("live-op")
+		Expect(err).ToNot(HaveOccurred())
+		Expect(live.Status).To(Equal("downloading"), "a recently-updated op must not be reaped")
+	})
+})

core/services/galleryop/managers.go

@@ -20,7 +20,7 @@ type BackendManager interface {
 	InstallBackend(ctx context.Context, op *ManagementOp[gallery.GalleryBackend, any], progressCb ProgressCallback) error
 	DeleteBackend(name string) error
 	ListBackends() (gallery.SystemBackends, error)
-	UpgradeBackend(ctx context.Context, name string, progressCb ProgressCallback) error
+	UpgradeBackend(ctx context.Context, opID, name string, progressCb ProgressCallback) error
 	CheckUpgrades(ctx context.Context) (map[string]gallery.UpgradeInfo, error)
 	// IsDistributed reports whether installs fan out across worker nodes.
 	// The HTTP layer uses this to refuse hardware-specific (non-meta) installs

core/services/galleryop/managers_local.go

@@ -96,7 +96,10 @@ func (b *LocalBackendManager) ListBackends() (gallery.SystemBackends, error) {
 	return gallery.ListSystemBackends(b.systemState)
 }
 
-func (b *LocalBackendManager) UpgradeBackend(ctx context.Context, name string, progressCb ProgressCallback) error {
+// UpgradeBackend ignores opID: a single-node install reports progress through
+// the local progressCb already; opID only matters for distributed per-node
+// streaming (see DistributedBackendManager.UpgradeBackend).
+func (b *LocalBackendManager) UpgradeBackend(ctx context.Context, _ string, name string, progressCb ProgressCallback) error {
 	return gallery.UpgradeBackend(ctx, b.systemState, b.modelLoader, b.backendGalleries, name, progressCb, b.requireBackendIntegrity)
 }
 

core/services/galleryop/opcache_evict_test.go

@@ -0,0 +1,107 @@
+package galleryop_test
+
+import (
+	"errors"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/config"
+	"github.com/mudler/LocalAI/core/services/galleryop"
+)
+
+// These specs reproduce the distributed "Reinstall spins forever" bug:
+// processingBackends (the UI spinner source) is built from OpCache.GetStatus,
+// which historically returned every cached op unconditionally. Cleanup only
+// happened when a client polled /api/backends/job/:uid, but the Manage-page
+// Reinstall/Upgrade buttons never poll, so a completed install stayed in
+// processingBackends forever. GetStatus must self-evict terminal ops.
+var _ = Describe("OpCache.GetStatus eviction", func() {
+	var (
+		svc   *galleryop.GalleryService
+		cache *galleryop.OpCache
+	)
+
+	BeforeEach(func() {
+		svc = galleryop.NewGalleryService(&config.ApplicationConfig{}, nil)
+		cache = galleryop.NewOpCache(svc)
+	})
+
+	It("keeps an op that is still processing", func() {
+		cache.SetBackend("llama-cpp", "uuid-1")
+		svc.UpdateStatus("uuid-1", &galleryop.OpStatus{Message: "processing backend: llama-cpp", Progress: 0})
+		processing, _ := cache.GetStatus()
+		Expect(processing).To(HaveKeyWithValue("llama-cpp", "uuid-1"))
+		Expect(cache.Exists("llama-cpp")).To(BeTrue())
+	})
+
+	It("evicts a completed op so it no longer shows as processing", func() {
+		cache.SetBackend("llama-cpp", "uuid-1")
+		svc.UpdateStatus("uuid-1", &galleryop.OpStatus{Processed: true, Progress: 100, Message: "completed"})
+		processing, _ := cache.GetStatus()
+		Expect(processing).NotTo(HaveKey("llama-cpp"))
+		Expect(cache.Exists("llama-cpp")).To(BeFalse())
+	})
+
+	It("keeps a failed op so the operations panel can surface the error and offer Dismiss", func() {
+		cache.SetBackend("piper", "uuid-2")
+		svc.UpdateStatus("uuid-2", &galleryop.OpStatus{Processed: true, Error: errors.New("boom")})
+		processing, _ := cache.GetStatus()
+		Expect(processing).To(HaveKeyWithValue("piper", "uuid-2"))
+		Expect(cache.Exists("piper")).To(BeTrue())
+	})
+
+	It("keeps the ErrWorkerStillInstalling soft-path op (worker still installing in background)", func() {
+		// Processed=true with no error but progress != 100 and a non-"completed"
+		// message: the worker timed out the NATS round-trip but is still installing.
+		// Evicting it would hide an install that may still fail; the reconciler
+		// confirms the real outcome later.
+		cache.SetBackend("vllm-development", "uuid-soft")
+		svc.UpdateStatus("uuid-soft", &galleryop.OpStatus{
+			Processed: true,
+			Message:   "backend vllm-development: worker still installing in background; reconciler will confirm completion",
+		})
+		processing, _ := cache.GetStatus()
+		Expect(processing).To(HaveKeyWithValue("vllm-development", "uuid-soft"))
+		Expect(cache.Exists("vllm-development")).To(BeTrue())
+	})
+
+	It("evicts a cancelled op", func() {
+		cache.SetBackend("vllm", "uuid-3")
+		svc.UpdateStatus("uuid-3", &galleryop.OpStatus{Processed: true, Cancelled: true, Message: "cancelled"})
+		processing, _ := cache.GetStatus()
+		Expect(processing).NotTo(HaveKey("vllm"))
+	})
+
+	It("does not evict an op with no status yet (queued)", func() {
+		cache.SetBackend("whisper", "uuid-4")
+		processing, taskTypes := cache.GetStatus()
+		Expect(processing).To(HaveKeyWithValue("whisper", "uuid-4"))
+		Expect(taskTypes).To(HaveKeyWithValue("whisper", "Waiting"))
+	})
+
+	// Regression guard: GetStatus is called concurrently by four HTTP handlers
+	// (~1s poll). An earlier version evicted by deleting from m.Map() — which
+	// returns the live internal map by reference — causing a fatal
+	// "concurrent map writes" crash. Run under -race; must not panic or race.
+	It("is safe under concurrent GetStatus + Set/complete", func() {
+		done := make(chan struct{})
+		go func() {
+			defer GinkgoRecover()
+			for i := 0; i < 2000; i++ {
+				_, _ = cache.GetStatus()
+			}
+			close(done)
+		}()
+		for i := 0; i < 2000; i++ {
+			id := "uuid-c"
+			cache.SetBackend("concurrent-backend", id)
+			// Half the time mark it completed so GetStatus evicts it.
+			if i%2 == 0 {
+				svc.UpdateStatus(id, &galleryop.OpStatus{Processed: true, Progress: 100, Message: "completed"})
+			}
+			_, _ = cache.GetStatus()
+		}
+		<-done
+	})
+})

core/services/galleryop/operation.go

@@ -408,12 +408,43 @@ func (m *OpCache) Exists(key string) bool {
 }
 
 func (m *OpCache) GetStatus() (map[string]string, map[string]string) {
-	processingModelsData := m.Map()
-
 	taskTypes := map[string]string{}
+	processingModelsData := map[string]string{}
 
-	for k, v := range processingModelsData {
+	// Iterate a snapshot (Keys() copies) and build a fresh result map. We must
+	// NOT delete from m.Map() during the range: Map() returns the live internal
+	// map by reference, so a bare delete here would be an unsynchronized write
+	// to a map four HTTP handlers read every ~1s — a concurrent-map-write crash.
+	// Collect evictions and apply them via the locked DeleteUUID after the loop.
+	var evict []string
+	for _, k := range m.status.Keys() {
+		v := m.status.Get(k)
+		if v == "" {
+			continue // raced with a concurrent Delete
+		}
 		status := m.galleryService.GetStatus(v)
+		// Terminal ops must not keep showing as "processing". Cleanup was
+		// previously only triggered by a client polling /api/backends/job/:uid,
+		// but the Manage-page Reinstall/Upgrade buttons never poll, so completed
+		// ops leaked into processingBackends forever and the card spun
+		// "reinstalling" indefinitely. Evict here on the list read (the UI always
+		// calls this). DeleteUUID broadcasts the eviction so peer replicas converge.
+		//
+		// We evict ONLY a clean success (progress 100 + "completed", matching the
+		// job-poll's historical delete condition) or a cancellation. Deliberately
+		// NOT evicted:
+		//   - failed ops (Error != nil): kept so /api/operations can surface the
+		//     error and offer Dismiss.
+		//   - the ErrWorkerStillInstalling soft-path (Processed=true, Error=nil,
+		//     progress != 100): the worker is still installing in the background
+		//     and the reconciler confirms the real outcome later — evicting it
+		//     would hide an install that may still fail.
+		if status != nil && status.Processed &&
+			((status.Progress == 100 && status.Message == "completed") || status.Cancelled) {
+			evict = append(evict, v)
+			continue
+		}
+		processingModelsData[k] = v
 		taskTypes[k] = "Installation"
 		if status != nil && status.Deletion {
 			taskTypes[k] = "Deletion"
@@ -422,6 +453,10 @@ func (m *OpCache) GetStatus() (map[string]string, map[string]string) {
 		}
 	}
 
+	for _, v := range evict {
+		m.DeleteUUID(v)
+	}
+
 	return processingModelsData, taskTypes
 }
 

core/services/galleryop/service.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"fmt"
 	"sync"
+	"time"
 
 	"github.com/mudler/LocalAI/core/config"
 	"github.com/mudler/LocalAI/core/gallery"
@@ -31,9 +32,9 @@ type GalleryService struct {
 	// natsClient is the wider MessagingClient (Publisher + subscribe methods)
 	// when wired by the distributed startup path; broadcastSubs holds the
 	// progress + cancel subscriptions opened by SubscribeBroadcasts.
-	natsClient     messaging.MessagingClient
-	galleryStore   *distributed.GalleryStore
-	broadcastSubs  []messaging.Subscription
+	natsClient    messaging.MessagingClient
+	galleryStore  *distributed.GalleryStore
+	broadcastSubs []messaging.Subscription
 
 	// OnBackendOpCompleted is fired after every successful install/upgrade/delete
 	// on the backend channel. The Application wires this to UpgradeChecker.TriggerCheck
@@ -274,6 +275,29 @@ func (g *GalleryService) GetAllStatus() map[string]*OpStatus {
 	return g.statuses
 }
 
+// ReapStaleOperations marks abandoned in-progress operations (pending/
+// downloading/processing) older than `age` as failed, so an op orphaned by a
+// replica that died mid-flight does not linger as "processing" forever. The
+// store's CleanStale runs once on startup; this exposes it for periodic
+// invocation (a post-startup orphan is otherwise not reaped until the next
+// restart). No-op when no gallery store is wired. Returns rows reaped.
+func (g *GalleryService) ReapStaleOperations(age time.Duration) (int64, error) {
+	g.Lock()
+	store := g.galleryStore
+	g.Unlock()
+	if store == nil {
+		return 0, nil
+	}
+	n, err := store.CleanStale(age)
+	if err != nil {
+		return 0, err
+	}
+	if n > 0 {
+		xlog.Info("Reaped stale gallery operations", "count", n)
+	}
+	return n, nil
+}
+
 // CancelOperation cancels an in-progress operation by its ID.
 //
 // In distributed mode the UI's cancel click may land on a different replica
@@ -295,6 +319,7 @@ func (g *GalleryService) CancelOperation(id string) error {
 	}
 
 	nc := g.natsClient
+	store := g.galleryStore
 
 	if !localExists && nc == nil {
 		g.Unlock()
@@ -315,6 +340,17 @@ func (g *GalleryService) CancelOperation(id string) error {
 	}
 	g.Unlock()
 
+	// Persist the terminal status so the cancel survives a restart. Without
+	// this the row stays in its active state and re-hydrates straight back into
+	// processingBackends on the next replica boot — the UI spins again on an op
+	// the admin already cancelled. The peer that broadcasts wins the write; a
+	// no-op when standalone (store nil).
+	if store != nil {
+		if err := store.Cancel(id); err != nil {
+			xlog.Warn("Failed to persist gallery operation cancellation", "op_id", id, "error", err)
+		}
+	}
+
 	// I/O and user-provided callback after Unlock — the cancel-wildcard
 	// subscriber loops back into applyCancel on this same replica, which
 	// would otherwise deadlock on g.Mutex.

core/services/messaging/subjects.go

@@ -194,6 +194,14 @@ type BackendUpgradeRequest struct {
 	// but the field lets future per-replica metadata (e.g. progress reporting
 	// scoped to a slot) ride the same wire without a v3 type.
 	ReplicaIndex int32 `json:"replica_index,omitempty"`
+	// OpID identifies the admin-side operation. When non-empty the worker
+	// publishes BackendInstallProgressEvent values to
+	// SubjectNodeBackendInstallProgress(nodeID, OpID) while the force-reinstall
+	// runs, so the master can stream per-node progress for upgrades exactly as
+	// it already does for installs (an upgrade IS a force-reinstall, so the
+	// install-progress subject is reused rather than minting a new one — no new
+	// NATS permission or rolling-update compat surface). Empty on legacy callers.
+	OpID string `json:"op_id,omitempty"`
 }
 
 // BackendUpgradeReply mirrors BackendInstallReply minus Address — upgrade does

core/services/nodes/inflight.go

@@ -157,3 +157,82 @@ func (c *InFlightTrackingClient) Rerank(ctx context.Context, in *pb.RerankReques
 	res, err := c.Backend.Rerank(ctx, in, opts...)
 	return res, c.reconcile(err)
 }
+
+func (c *InFlightTrackingClient) VAD(ctx context.Context, in *pb.VADRequest, opts ...ggrpc.CallOption) (*pb.VADResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.VAD(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) Diarize(ctx context.Context, in *pb.DiarizeRequest, opts ...ggrpc.CallOption) (*pb.DiarizeResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.Diarize(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) FaceVerify(ctx context.Context, in *pb.FaceVerifyRequest, opts ...ggrpc.CallOption) (*pb.FaceVerifyResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.FaceVerify(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) FaceAnalyze(ctx context.Context, in *pb.FaceAnalyzeRequest, opts ...ggrpc.CallOption) (*pb.FaceAnalyzeResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.FaceAnalyze(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) VoiceVerify(ctx context.Context, in *pb.VoiceVerifyRequest, opts ...ggrpc.CallOption) (*pb.VoiceVerifyResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.VoiceVerify(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) VoiceAnalyze(ctx context.Context, in *pb.VoiceAnalyzeRequest, opts ...ggrpc.CallOption) (*pb.VoiceAnalyzeResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.VoiceAnalyze(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) VoiceEmbed(ctx context.Context, in *pb.VoiceEmbedRequest, opts ...ggrpc.CallOption) (*pb.VoiceEmbedResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.VoiceEmbed(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) TokenClassify(ctx context.Context, in *pb.TokenClassifyRequest, opts ...ggrpc.CallOption) (*pb.TokenClassifyResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.TokenClassify(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) Score(ctx context.Context, in *pb.ScoreRequest, opts ...ggrpc.CallOption) (*pb.ScoreResponse, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.Score(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) AudioEncode(ctx context.Context, in *pb.AudioEncodeRequest, opts ...ggrpc.CallOption) (*pb.AudioEncodeResult, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.AudioEncode(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) AudioDecode(ctx context.Context, in *pb.AudioDecodeRequest, opts ...ggrpc.CallOption) (*pb.AudioDecodeResult, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.AudioDecode(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+func (c *InFlightTrackingClient) AudioTransform(ctx context.Context, in *pb.AudioTransformRequest, opts ...ggrpc.CallOption) (*pb.AudioTransformResult, error) {
+	defer c.track(ctx)()
+	res, err := c.Backend.AudioTransform(ctx, in, opts...)
+	return res, c.reconcile(err)
+}
+
+// AudioTransformStream, AudioToAudioStream and Forward are deliberately left as
+// embedded passthrough: they return a stream client and the inference spans the
+// stream's lifetime, not the constructor call. Wrapping the constructor with
+// track() would increment and immediately decrement (and fire onFirstComplete)
+// before any audio flows. Tracking those correctly needs the done() func tied to
+// stream close, which the current Backend interface doesn't surface here.

core/services/nodes/inflight_test.go

@@ -304,6 +304,105 @@ var _ = Describe("InFlightTrackingClient", func() {
 		})
 	})
 
+	Describe("non-LLM inference methods track in-flight", func() {
+		// silero-vad and friends only ever expose a single non-Predict method.
+		// If that method isn't wrapped, the load-time reservation released by
+		// onFirstComplete never fires and in-flight is stuck at 1 forever.
+		assertTracked := func(call func() error) {
+			var firstFired int
+			client.OnFirstComplete(func() { firstFired++ })
+			err := call()
+			Expect(err).ToNot(HaveOccurred())
+			Expect(tracker.increments).To(Equal(1), "method must increment in-flight")
+			Expect(tracker.decrements).To(Equal(1), "method must decrement in-flight")
+			Expect(firstFired).To(Equal(1), "method must release the load-time reservation")
+		}
+
+		It("VAD", func() {
+			assertTracked(func() error {
+				_, err := client.VAD(context.Background(), &pb.VADRequest{})
+				return err
+			})
+		})
+
+		It("Diarize", func() {
+			assertTracked(func() error {
+				_, err := client.Diarize(context.Background(), &pb.DiarizeRequest{})
+				return err
+			})
+		})
+
+		It("VoiceVerify", func() {
+			assertTracked(func() error {
+				_, err := client.VoiceVerify(context.Background(), &pb.VoiceVerifyRequest{})
+				return err
+			})
+		})
+
+		It("VoiceAnalyze", func() {
+			assertTracked(func() error {
+				_, err := client.VoiceAnalyze(context.Background(), &pb.VoiceAnalyzeRequest{})
+				return err
+			})
+		})
+
+		It("VoiceEmbed", func() {
+			assertTracked(func() error {
+				_, err := client.VoiceEmbed(context.Background(), &pb.VoiceEmbedRequest{})
+				return err
+			})
+		})
+
+		It("FaceVerify", func() {
+			assertTracked(func() error {
+				_, err := client.FaceVerify(context.Background(), &pb.FaceVerifyRequest{})
+				return err
+			})
+		})
+
+		It("FaceAnalyze", func() {
+			assertTracked(func() error {
+				_, err := client.FaceAnalyze(context.Background(), &pb.FaceAnalyzeRequest{})
+				return err
+			})
+		})
+
+		It("TokenClassify", func() {
+			assertTracked(func() error {
+				_, err := client.TokenClassify(context.Background(), &pb.TokenClassifyRequest{})
+				return err
+			})
+		})
+
+		It("Score", func() {
+			assertTracked(func() error {
+				_, err := client.Score(context.Background(), &pb.ScoreRequest{})
+				return err
+			})
+		})
+
+		It("AudioEncode", func() {
+			assertTracked(func() error {
+				_, err := client.AudioEncode(context.Background(), &pb.AudioEncodeRequest{})
+				return err
+			})
+		})
+
+		It("AudioDecode", func() {
+			assertTracked(func() error {
+				_, err := client.AudioDecode(context.Background(), &pb.AudioDecodeRequest{})
+				return err
+			})
+		})
+
+		It("AudioTransform", func() {
+			assertTracked(func() error {
+				_, err := client.AudioTransform(context.Background(), &pb.AudioTransformRequest{})
+				return err
+			})
+		})
+	})
+
 	Describe("stale model reload (self-heal)", func() {
 		It("removes the replica when the backend reports the model is not loaded", func() {
 			backend.predictErr = fmt.Errorf("parakeet-cpp: model not loaded")

core/services/nodes/managers_distributed.go

@@ -533,7 +533,7 @@ func (d *DistributedBackendManager) InstallBackend(ctx context.Context, op *gall
 // backend.upgrade, we try the legacy backend.install Force=true path so a
 // new master + old worker still converges. Drop the fallback once every
 // worker in the fleet is on 2026-05-08 or newer.
-func (d *DistributedBackendManager) UpgradeBackend(ctx context.Context, name string, progressCb galleryop.ProgressCallback) error {
+func (d *DistributedBackendManager) UpgradeBackend(ctx context.Context, opID, name string, progressCb galleryop.ProgressCallback) error {
 	galleriesJSON, _ := json.Marshal(d.backendGalleries)
 
 	installed, err := d.ListBackends()
@@ -549,17 +549,39 @@ func (d *DistributedBackendManager) UpgradeBackend(ctx context.Context, name str
 		targetNodeIDs[n.NodeID] = true
 	}
 
-	// Empty opID: the caller (galleryop) doesn't thread an op ID into
-	// UpgradeBackend today, so we can't tag per-node sink writes with the
-	// right OpStatus key. Until the upgrade path takes a ManagementOp the
-	// way InstallBackend does, the sink stays no-op here.
-	result, err := d.enqueueAndDrainBackendOp(ctx, "", OpBackendUpgrade, name, galleriesJSON, targetNodeIDs, func(node BackendNode) error {
-		reply, err := d.adapter.UpgradeBackend(node.ID, name, string(galleriesJSON), "", "", "", 0)
+	result, err := d.enqueueAndDrainBackendOp(ctx, opID, OpBackendUpgrade, name, galleriesJSON, targetNodeIDs, func(node BackendNode) error {
+		// Per-node progress sink: fan each worker download tick into the legacy
+		// single-bar progressCb and the per-node OpStatus.Nodes view, exactly as
+		// InstallBackend does. Defined per-node so each closure captures its own
+		// node.Name. Without this an upgrade blocks opaque at progress 0 for the
+		// whole 15m round-trip (the original "reinstalling but nothing happens").
+		onProgress := func(ev messaging.BackendInstallProgressEvent) {
+			if progressCb != nil {
+				progressCb(ev.FileName, ev.Current, ev.Total, ev.Percentage)
+			}
+			if d.progressSink != nil && opID != "" {
+				d.progressSink.UpdateNodeProgress(opID, ev.NodeID, galleryop.NodeProgress{
+					NodeID:     ev.NodeID,
+					NodeName:   node.Name,
+					Status:     galleryop.NodeStatusDownloading,
+					FileName:   ev.FileName,
+					Current:    ev.Current,
+					Total:      ev.Total,
+					Percentage: ev.Percentage,
+					Phase:      ev.Phase,
+				})
+			}
+		}
+		var onProgressArg func(messaging.BackendInstallProgressEvent)
+		if progressCb != nil || d.progressSink != nil {
+			onProgressArg = onProgress
+		}
+		reply, err := d.adapter.UpgradeBackend(node.ID, name, string(galleriesJSON), "", "", "", 0, opID, onProgressArg)
 		if err != nil {
 			// Rolling-update fallback: an older worker doesn't know
 			// backend.upgrade. Try the legacy install-with-force path.
 			if errors.Is(err, nats.ErrNoResponders) {
-				instReply, instErr := d.adapter.installWithForceFallback(node.ID, name, string(galleriesJSON), "", "", "", 0)
+				instReply, instErr := d.adapter.installWithForceFallback(node.ID, name, string(galleriesJSON), "", "", "", 0, opID, onProgressArg)
 				if instErr != nil {
 					return instErr
 				}

core/services/nodes/managers_distributed_test.go

@@ -317,7 +317,7 @@ func (stubLocalBackendManager) DeleteBackend(_ string) error { return gallery.Er
 func (stubLocalBackendManager) ListBackends() (gallery.SystemBackends, error) {
 	return gallery.SystemBackends{}, nil
 }
-func (stubLocalBackendManager) UpgradeBackend(_ context.Context, _ string, _ galleryop.ProgressCallback) error {
+func (stubLocalBackendManager) UpgradeBackend(_ context.Context, _ string, _ string, _ galleryop.ProgressCallback) error {
 	return nil
 }
 func (stubLocalBackendManager) CheckUpgrades(_ context.Context) (map[string]gallery.UpgradeInfo, error) {
@@ -782,7 +782,7 @@ var _ = Describe("DistributedBackendManager", func() {
 				mc.scriptReply(messaging.SubjectNodeBackendUpgrade(n2.ID),
 					messaging.BackendUpgradeReply{Success: false, Error: "registry unauthorized"})
 
-				err := mgr.UpgradeBackend(ctx, "vllm-development", nil)
+				err := mgr.UpgradeBackend(ctx, "", "vllm-development", nil)
 				Expect(err).To(HaveOccurred())
 				Expect(err.Error()).To(ContainSubstring("worker-a"))
 				Expect(err.Error()).To(ContainSubstring("image manifest not found"))
@@ -797,7 +797,7 @@ var _ = Describe("DistributedBackendManager", func() {
 				scriptInstalled("vllm-development", n1.ID)
 				mc.scriptReply(messaging.SubjectNodeBackendUpgrade(n1.ID),
 					messaging.BackendUpgradeReply{Success: true})
-				Expect(mgr.UpgradeBackend(ctx, "vllm-development", nil)).To(Succeed())
+				Expect(mgr.UpgradeBackend(ctx, "", "vllm-development", nil)).To(Succeed())
 			})
 		})
 
@@ -819,7 +819,7 @@ var _ = Describe("DistributedBackendManager", func() {
 				// if the manager attempts it, the scripted-client default returns
 				// fakeNoRespondersErr and the assertion below fails loudly.
 
-				Expect(mgr.UpgradeBackend(ctx, "cpu-insightface-development", nil)).To(Succeed())
+				Expect(mgr.UpgradeBackend(ctx, "", "cpu-insightface-development", nil)).To(Succeed())
 
 				mc.mu.Lock()
 				defer mc.mu.Unlock()
@@ -835,7 +835,7 @@ var _ = Describe("DistributedBackendManager", func() {
 				n1 := registerHealthyBackend("worker-a", "10.0.0.1:50051")
 				scriptNoBackends(n1.ID)
 
-				err := mgr.UpgradeBackend(ctx, "vllm-development", nil)
+				err := mgr.UpgradeBackend(ctx, "", "vllm-development", nil)
 				Expect(err).To(HaveOccurred())
 				Expect(err.Error()).To(ContainSubstring("not installed on any node"))
 
@@ -865,7 +865,7 @@ var _ = Describe("DistributedBackendManager", func() {
 					func(req messaging.BackendInstallRequest) bool { return req.Force },
 					messaging.BackendInstallReply{Success: true, Address: "10.0.0.1:50100"})
 
-				Expect(mgr.UpgradeBackend(ctx, "vllm-development", nil)).To(Succeed())
+				Expect(mgr.UpgradeBackend(ctx, "", "vllm-development", nil)).To(Succeed())
 			})
 
 			It("returns the upgrade error when it is not ErrNoResponders", func() {
@@ -875,7 +875,7 @@ var _ = Describe("DistributedBackendManager", func() {
 				mc.scriptReply(messaging.SubjectNodeBackendUpgrade(n.ID),
 					messaging.BackendUpgradeReply{Success: false, Error: "disk full"})
 
-				err := mgr.UpgradeBackend(ctx, "vllm-development", nil)
+				err := mgr.UpgradeBackend(ctx, "", "vllm-development", nil)
 				Expect(err).To(HaveOccurred())
 				Expect(err.Error()).To(ContainSubstring("disk full"))
 			})

core/services/nodes/pending_op_cleanup_test.go

@@ -0,0 +1,135 @@
+package nodes
+
+import (
+	"context"
+	"runtime"
+	"time"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+
+	"github.com/mudler/LocalAI/core/services/testutil"
+)
+
+// These specs reproduce the distributed "pending ops behind dead nodes leak
+// forever" bug. ListDuePendingBackendOps only returns rows whose node is
+// StatusHealthy, so an op queued against a node that goes offline (heartbeat
+// stale) or draining (admin action) is never retried, never aged out, and
+// never deleted. On a live cluster these rows sat at attempts=0 indefinitely
+// and kept the UI operation alive. DeleteStalePendingBackendOps garbage-collects
+// them: draining nodes immediately (models already purged), offline nodes only
+// after a grace window so a brief heartbeat blip does not nuke in-flight work.
+var _ = Describe("DeleteStalePendingBackendOps", func() {
+	var (
+		registry *NodeRegistry
+		ctx      context.Context
+	)
+
+	BeforeEach(func() {
+		if runtime.GOOS == "darwin" {
+			Skip("testcontainers requires Docker, not available on macOS CI")
+		}
+		db := testutil.SetupTestDB()
+		var err error
+		registry, err = NewNodeRegistry(db)
+		Expect(err).ToNot(HaveOccurred())
+		ctx = context.Background()
+	})
+
+	// registerBackend registers an auto-approved backend node and returns its ID.
+	registerBackend := func(name, address string) string {
+		node := &BackendNode{Name: name, NodeType: NodeTypeBackend, Address: address}
+		Expect(registry.Register(ctx, node, true)).To(Succeed())
+		fetched, err := registry.GetByName(ctx, name)
+		Expect(err).ToNot(HaveOccurred())
+		return fetched.ID
+	}
+
+	// setHeartbeat forces a node's last_heartbeat (Register/MarkOffline leave it
+	// at "now"; we age it to simulate a node that went silent a while ago).
+	setHeartbeat := func(nodeID string, t time.Time) {
+		Expect(registry.db.WithContext(ctx).Model(&BackendNode{}).
+			Where("id = ?", nodeID).
+			Update("last_heartbeat", t).Error).To(Succeed())
+	}
+
+	pendingCountFor := func(nodeID string) int64 {
+		var n int64
+		Expect(registry.db.WithContext(ctx).Model(&PendingBackendOp{}).
+			Where("node_id = ?", nodeID).Count(&n).Error).To(Succeed())
+		return n
+	}
+
+	It("clears ops behind an offline node whose heartbeat is past the grace window", func() {
+		dead := registerBackend("nvidia-thor", "10.0.0.9:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, dead, "llama-cpp-development", OpBackendInstall, nil)).To(Succeed())
+		Expect(registry.MarkOffline(ctx, dead)).To(Succeed())
+		setHeartbeat(dead, time.Now().Add(-1*time.Hour))
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(1)))
+		Expect(pendingCountFor(dead)).To(Equal(int64(0)))
+	})
+
+	It("clears ops behind a draining node immediately, even with a fresh heartbeat", func() {
+		// Mirrors the live mac-mini-m4 case: draining but still heartbeating.
+		drain := registerBackend("mac-mini-m4", "10.0.0.3:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, drain, "llama-cpp-development", OpBackendInstall, nil)).To(Succeed())
+		Expect(registry.MarkDraining(ctx, drain)).To(Succeed())
+		setHeartbeat(drain, time.Now()) // fresh heartbeat
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(1)))
+		Expect(pendingCountFor(drain)).To(Equal(int64(0)))
+	})
+
+	It("clears ops behind an unhealthy node with a stale heartbeat (never ages to offline)", func() {
+		// A node marked unhealthy on a NATS ErrNoResponders never transitions to
+		// offline, so its ops must be reaped via the same stale-heartbeat path.
+		sick := registerBackend("agx-orin-sick", "10.0.0.7:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, sick, "llama-cpp-development", OpBackendUpgrade, nil)).To(Succeed())
+		Expect(registry.MarkUnhealthy(ctx, sick)).To(Succeed())
+		setHeartbeat(sick, time.Now().Add(-1*time.Hour))
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(1)))
+		Expect(pendingCountFor(sick)).To(Equal(int64(0)))
+	})
+
+	It("keeps ops behind an unhealthy node that is still heartbeating (recovering)", func() {
+		recovering := registerBackend("agx-orin-flap", "10.0.0.8:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, recovering, "llama-cpp-development", OpBackendUpgrade, nil)).To(Succeed())
+		Expect(registry.MarkUnhealthy(ctx, recovering)).To(Succeed())
+		setHeartbeat(recovering, time.Now()) // fresh heartbeat → recovering
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(0)))
+		Expect(pendingCountFor(recovering)).To(Equal(int64(1)))
+	})
+
+	It("keeps ops behind a node that only just went offline (within grace)", func() {
+		blip := registerBackend("agx-orin", "10.0.0.4:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, blip, "parakeet-cpp-development", OpBackendInstall, nil)).To(Succeed())
+		Expect(registry.MarkOffline(ctx, blip)).To(Succeed())
+		setHeartbeat(blip, time.Now().Add(-1*time.Minute)) // gone only 1m, grace 10m
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(0)))
+		Expect(pendingCountFor(blip)).To(Equal(int64(1)))
+	})
+
+	It("keeps ops behind a healthy node", func() {
+		healthy := registerBackend("dgx-spark", "10.0.0.1:50051")
+		Expect(registry.UpsertPendingBackendOp(ctx, healthy, "llama-cpp-development", OpBackendUpgrade, nil)).To(Succeed())
+
+		removed, err := registry.DeleteStalePendingBackendOps(ctx, 10*time.Minute)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(removed).To(Equal(int64(0)))
+		Expect(pendingCountFor(healthy)).To(Equal(int64(1)))
+	})
+})

core/services/nodes/reconciler.go

@@ -189,6 +189,13 @@ func (rc *ReplicaReconciler) reconcileState(ctx context.Context) {
 // passed on nodes that are currently healthy. On success the row is deleted;
 // on failure attempts++ and next_retry_at moves out via exponential backoff.
 func (rc *ReplicaReconciler) drainPendingBackendOps(ctx context.Context) {
+	// Garbage-collect ops behind nodes that went offline/draining. These are
+	// invisible to ListDuePendingBackendOps (which filters status=healthy), so
+	// without this sweep they leak forever and keep the UI operation spinning.
+	if _, err := rc.registry.DeleteStalePendingBackendOps(ctx, stalePendingBackendOpGrace); err != nil {
+		xlog.Warn("Reconciler: failed to clear stale pending backend ops", "error", err)
+	}
+
 	ops, err := rc.registry.ListDuePendingBackendOps(ctx)
 	if err != nil {
 		xlog.Warn("Reconciler: failed to list pending backend ops", "error", err)
@@ -223,10 +230,13 @@ func (rc *ReplicaReconciler) drainPendingBackendOps(ctx context.Context) {
 			// the same worker. Falls back to the legacy backend.install
 			// Force=true path on nats.ErrNoResponders for old workers that
 			// don't subscribe to backend.upgrade yet (rolling-update window).
-			reply, err := rc.adapter.UpgradeBackend(op.NodeID, op.Backend, string(op.Galleries), "", "", "", 0)
+			// Reconciler retries are background reconciliation with no live
+			// admin watching a progress bar, so opID/onProgress are empty —
+			// the adapter skips the progress subscription entirely.
+			reply, err := rc.adapter.UpgradeBackend(op.NodeID, op.Backend, string(op.Galleries), "", "", "", 0, "", nil)
 			if err != nil {
 				if errors.Is(err, nats.ErrNoResponders) {
-					instReply, instErr := rc.adapter.installWithForceFallback(op.NodeID, op.Backend, string(op.Galleries), "", "", "", 0)
+					instReply, instErr := rc.adapter.installWithForceFallback(op.NodeID, op.Backend, string(op.Galleries), "", "", "", 0, "", nil)
 					if instErr != nil {
 						applyErr = instErr
 					} else if !instReply.Success {
@@ -293,6 +303,13 @@ func (rc *ReplicaReconciler) drainPendingBackendOps(ctx context.Context) {
 // amount of further retrying will help.
 const maxPendingBackendOpAttempts = 10
 
+// stalePendingBackendOpGrace is how long a node may be offline before its
+// pending backend ops are garbage-collected. Draining nodes are cleared
+// immediately regardless of this window (see DeleteStalePendingBackendOps).
+// ListDuePendingBackendOps never surfaces ops behind non-healthy nodes, so
+// without this sweep they would leak forever and keep the UI op spinning.
+const stalePendingBackendOpGrace = 15 * time.Minute
+
 // probeLoadedModels gRPC-health-checks model addresses that the DB says are
 // loaded. If a model's backend process is gone (OOM, crash, manual restart)
 // we remove the row so ghosts don't linger. Only probes rows older than

core/services/nodes/registry.go

@@ -1776,6 +1776,38 @@ func (r *NodeRegistry) DeletePendingBackendOp(ctx context.Context, id uint) erro
 	return nil
 }
 
+// DeleteStalePendingBackendOps garbage-collects pending backend ops whose target
+// node can never drain them. ListDuePendingBackendOps only returns rows behind a
+// StatusHealthy node, so ops behind a node that went offline or draining are
+// otherwise never retried, aged out, or deleted — they leak forever and keep the
+// UI operation spinning. Draining nodes are cleared immediately (an explicit
+// admin action; their model rows are already purged). Offline nodes are cleared
+// only once their last heartbeat is older than `grace`, so a brief heartbeat blip
+// does not nuke an install that is still legitimately in flight. Returns the
+// number of rows deleted.
+func (r *NodeRegistry) DeleteStalePendingBackendOps(ctx context.Context, grace time.Duration) (int64, error) {
+	cutoff := time.Now().Add(-grace)
+	// Draining nodes are cleared immediately (admin action; model rows already
+	// purged). Offline AND unhealthy nodes are cleared only once their heartbeat
+	// is older than the grace window: a node marked unhealthy on a NATS
+	// ErrNoResponders never transitions to offline (health.go skips re-marking
+	// it), so without including unhealthy here its ops would leak exactly like
+	// the offline case. A node with a fresh heartbeat (last_heartbeat > cutoff)
+	// is recovering and keeps its op for retry.
+	res := r.db.WithContext(ctx).
+		Where(`node_id IN (SELECT id FROM backend_nodes WHERE status = ?)
+			OR node_id IN (SELECT id FROM backend_nodes WHERE status IN ? AND last_heartbeat <= ?)`,
+			StatusDraining, []string{StatusOffline, StatusUnhealthy}, cutoff).
+		Delete(&PendingBackendOp{})
+	if res.Error != nil {
+		return 0, fmt.Errorf("deleting stale pending backend ops: %w", res.Error)
+	}
+	if res.RowsAffected > 0 {
+		xlog.Info("Cleared pending backend ops behind non-healthy nodes", "deleted", res.RowsAffected)
+	}
+	return res.RowsAffected, nil
+}
+
 // RecordPendingBackendOpFailure bumps Attempts, captures the error, and
 // pushes NextRetryAt out with exponential backoff capped at 15 minutes.
 func (r *NodeRegistry) RecordPendingBackendOpFailure(ctx context.Context, id uint, errMsg string) error {

core/services/nodes/router_test.go

@@ -365,7 +365,7 @@ func (f *fakeUnloader) InstallBackend(nodeID, backend, modelID, _, _, _, _ strin
 	return f.installReply, f.installErr
 }
 
-func (f *fakeUnloader) UpgradeBackend(nodeID, backend, _, _, _, _ string, replica int) (*messaging.BackendUpgradeReply, error) {
+func (f *fakeUnloader) UpgradeBackend(nodeID, backend, _, _, _, _ string, replica int, _ string, _ func(messaging.BackendInstallProgressEvent)) (*messaging.BackendUpgradeReply, error) {
 	f.mu.Lock()
 	f.upgradeCalls = append(f.upgradeCalls, upgradeCall{nodeID, backend, replica})
 	f.mu.Unlock()

core/services/nodes/unloader.go

@@ -35,7 +35,7 @@ type backendStopRequest struct {
 // backend.upgrade subject.
 type NodeCommandSender interface {
 	InstallBackend(nodeID, backendType, modelID, galleriesJSON, uri, name, alias string, replicaIndex int, opID string, onProgress func(messaging.BackendInstallProgressEvent)) (*messaging.BackendInstallReply, error)
-	UpgradeBackend(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int) (*messaging.BackendUpgradeReply, error)
+	UpgradeBackend(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int, opID string, onProgress func(messaging.BackendInstallProgressEvent)) (*messaging.BackendUpgradeReply, error)
 	DeleteBackend(nodeID, backendName string) (*messaging.BackendDeleteReply, error)
 	ListBackends(nodeID string) (*messaging.BackendListReply, error)
 	StopBackend(nodeID, backend string) error
@@ -127,38 +127,8 @@ func (a *RemoteUnloaderAdapter) InstallBackend(
 	xlog.Info("Sending NATS backend.install", "nodeID", nodeID, "backend", backendType, "modelID", modelID, "replica", replicaIndex, "opID", opID)
 
 	// Subscribe to the per-op progress subject BEFORE publishing the install
-	// request so we don't miss early events. When onProgress is nil OR opID
-	// is empty (the reconciler-driven retry path), skip subscription entirely:
-	// silent installs cost nothing extra.
-	var sub messaging.Subscription
-	if onProgress != nil && opID != "" {
-		progressSubject := messaging.SubjectNodeBackendInstallProgress(nodeID, opID)
-		s, subErr := a.nats.Subscribe(progressSubject, func(raw []byte) {
-			var ev messaging.BackendInstallProgressEvent
-			if err := json.Unmarshal(raw, &ev); err != nil {
-				xlog.Debug("malformed install progress event", "subject", progressSubject, "error", err)
-				return
-			}
-			// Goroutine guard: a slow onProgress callback must not stall
-			// the NATS reader thread.
-			//
-			// NOTE: events spawn one goroutine each, so ordering at the
-			// consumer is best-effort. In practice the worker debounces to
-			// ~250ms which is far larger than goroutine scheduling jitter,
-			// so reordering is rare. The worker's final Flush() event is
-			// intended to win as the terminal tick. A future hardening pass
-			// could add a Seq uint64 field to BackendInstallProgressEvent
-			// and drop stale-by-seq at the bridge if reordering becomes a
-			// real UX issue.
-			go onProgress(ev)
-		})
-		if subErr != nil {
-			xlog.Warn("Failed to subscribe to install progress subject; proceeding without progress streaming",
-				"subject", progressSubject, "error", subErr)
-		} else {
-			sub = s
-		}
-	}
+	// request so we don't miss early events.
+	sub := a.subscribeProgress(nodeID, opID, onProgress)
 
 	reply, err := messaging.RequestJSON[messaging.BackendInstallRequest, messaging.BackendInstallReply](a.nats, subject, messaging.BackendInstallRequest{
 		Backend:          backendType,
@@ -182,18 +152,58 @@ func (a *RemoteUnloaderAdapter) InstallBackend(
 	return reply, err
 }
 
+// subscribeProgress subscribes to the per-op backend-install progress subject
+// so the master can stream per-node download ticks while a worker installs or
+// upgrades. Returns nil (and subscribes to nothing) when onProgress is nil or
+// opID is empty — the reconciler-driven retry path and legacy callers stay
+// silent at no cost. Shared by InstallBackend, UpgradeBackend, and the legacy
+// force-install fallback: an upgrade is a force-reinstall, so it reuses the
+// install-progress subject rather than minting a new one (no new NATS
+// permission, no new rolling-update compat surface). Caller must Unsubscribe
+// the returned subscription after the request completes.
+func (a *RemoteUnloaderAdapter) subscribeProgress(nodeID, opID string, onProgress func(messaging.BackendInstallProgressEvent)) messaging.Subscription {
+	if onProgress == nil || opID == "" {
+		return nil
+	}
+	progressSubject := messaging.SubjectNodeBackendInstallProgress(nodeID, opID)
+	s, subErr := a.nats.Subscribe(progressSubject, func(raw []byte) {
+		var ev messaging.BackendInstallProgressEvent
+		if err := json.Unmarshal(raw, &ev); err != nil {
+			xlog.Debug("malformed backend progress event", "subject", progressSubject, "error", err)
+			return
+		}
+		// Goroutine guard: a slow onProgress callback must not stall the NATS
+		// reader thread. Events spawn one goroutine each, so ordering at the
+		// consumer is best-effort; the worker debounces to ~250ms which dwarfs
+		// goroutine scheduling jitter, and its final Flush() is the terminal tick.
+		go onProgress(ev)
+	})
+	if subErr != nil {
+		xlog.Warn("Failed to subscribe to backend progress subject; proceeding without progress streaming",
+			"subject", progressSubject, "error", subErr)
+		return nil
+	}
+	return s
+}
+
 // UpgradeBackend sends a backend.upgrade request-reply to a worker node.
 // The worker stops every live process for this backend, force-reinstalls
 // from the gallery (overwriting the on-disk artifact), and replies. The
 // next routine InstallBackend call spawns a fresh process with the new
 // binary - upgrade itself does not start a process.
 //
+// When opID is non-empty and onProgress is set, the master subscribes to the
+// per-op progress subject before firing the request so a long force-reinstall
+// streams per-node download ticks instead of blocking opaque at progress 0.
+//
 // Timeout: configured via DistributedConfig.BackendUpgradeTimeoutOrDefault
 // (default 15m). Real-world worst case observed: 8-10 minutes for large
 // CUDA-l4t backend images on Jetson over WiFi.
-func (a *RemoteUnloaderAdapter) UpgradeBackend(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int) (*messaging.BackendUpgradeReply, error) {
+func (a *RemoteUnloaderAdapter) UpgradeBackend(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int, opID string, onProgress func(messaging.BackendInstallProgressEvent)) (*messaging.BackendUpgradeReply, error) {
 	subject := messaging.SubjectNodeBackendUpgrade(nodeID)
-	xlog.Info("Sending NATS backend.upgrade", "nodeID", nodeID, "backend", backendType, "replica", replicaIndex)
+	xlog.Info("Sending NATS backend.upgrade", "nodeID", nodeID, "backend", backendType, "replica", replicaIndex, "opID", opID)
+
+	sub := a.subscribeProgress(nodeID, opID, onProgress)
 
 	reply, err := messaging.RequestJSON[messaging.BackendUpgradeRequest, messaging.BackendUpgradeReply](a.nats, subject, messaging.BackendUpgradeRequest{
 		Backend:          backendType,
@@ -202,7 +212,13 @@ func (a *RemoteUnloaderAdapter) UpgradeBackend(nodeID, backendType, galleriesJSO
 		Name:             name,
 		Alias:            alias,
 		ReplicaIndex:     int32(replicaIndex),
+		OpID:             opID,
 	}, a.upgradeTimeout)
+
+	if sub != nil {
+		_ = sub.Unsubscribe()
+	}
+
 	if err != nil && isNATSTimeout(err) {
 		return nil, fmt.Errorf("%w (subject=%s nodeID=%s backend=%s): %v",
 			galleryop.ErrWorkerStillInstalling, subject, nodeID, backendType, err)
@@ -216,10 +232,12 @@ func (a *RemoteUnloaderAdapter) UpgradeBackend(nodeID, backendType, galleriesJSO
 // doesn't subscribe to the new subject). It re-fires the legacy
 // backend.install with Force=true. Drop this once every worker is on
 // 2026-05-08 or newer.
-func (a *RemoteUnloaderAdapter) installWithForceFallback(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int) (*messaging.BackendInstallReply, error) {
+func (a *RemoteUnloaderAdapter) installWithForceFallback(nodeID, backendType, galleriesJSON, uri, name, alias string, replicaIndex int, opID string, onProgress func(messaging.BackendInstallProgressEvent)) (*messaging.BackendInstallReply, error) {
 	subject := messaging.SubjectNodeBackendInstall(nodeID)
 	xlog.Warn("Falling back to legacy backend.install Force=true (old worker)", "nodeID", nodeID, "backend", backendType)
 
+	sub := a.subscribeProgress(nodeID, opID, onProgress)
+
 	reply, err := messaging.RequestJSON[messaging.BackendInstallRequest, messaging.BackendInstallReply](a.nats, subject, messaging.BackendInstallRequest{
 		Backend:          backendType,
 		BackendGalleries: galleriesJSON,
@@ -228,7 +246,13 @@ func (a *RemoteUnloaderAdapter) installWithForceFallback(nodeID, backendType, ga
 		Alias:            alias,
 		ReplicaIndex:     int32(replicaIndex),
 		Force:            true,
+		OpID:             opID,
 	}, a.upgradeTimeout)
+
+	if sub != nil {
+		_ = sub.Unsubscribe()
+	}
+
 	if err != nil && isNATSTimeout(err) {
 		return nil, fmt.Errorf("%w (subject=%s nodeID=%s backend=%s): %v",
 			galleryop.ErrWorkerStillInstalling, subject, nodeID, backendType, err)

core/services/nodes/unloader_test.go

@@ -282,7 +282,7 @@ var _ = Describe("RemoteUnloaderAdapter timeout configuration", func() {
 		mc.scriptReply(messaging.SubjectNodeBackendUpgrade("n1"), messaging.BackendUpgradeReply{Success: true})
 		adapter := NewRemoteUnloaderAdapter(nil, mc, 7*time.Minute, 11*time.Minute)
 
-		_, err := adapter.UpgradeBackend("n1", "llama-cpp", "[]", "", "", "", 0)
+		_, err := adapter.UpgradeBackend("n1", "llama-cpp", "[]", "", "", "", 0, "", nil)
 		Expect(err).ToNot(HaveOccurred())
 
 		Expect(mc.calls).To(HaveLen(1))

core/services/nodes/unloader_upgrade_test.go

@@ -1,6 +1,7 @@
 package nodes
 
 import (
+	"sync"
 	"time"
 
 	. "github.com/onsi/ginkgo/v2"
@@ -18,7 +19,7 @@ var _ = Describe("RemoteUnloaderAdapter.UpgradeBackend", func() {
 			messaging.BackendUpgradeReply{Success: true})
 
 		adapter := NewRemoteUnloaderAdapter(nil, mc, 3*time.Minute, 15*time.Minute)
-		reply, err := adapter.UpgradeBackend(nodeID, "llama-cpp", `[{"name":"x"}]`, "", "", "", 0)
+		reply, err := adapter.UpgradeBackend(nodeID, "llama-cpp", `[{"name":"x"}]`, "", "", "", 0, "", nil)
 		Expect(err).ToNot(HaveOccurred())
 		Expect(reply.Success).To(BeTrue())
 	})
@@ -27,7 +28,55 @@ var _ = Describe("RemoteUnloaderAdapter.UpgradeBackend", func() {
 		mc := newScriptedMessagingClient() // unscripted subject => fakeNoRespondersErr by harness convention
 
 		adapter := NewRemoteUnloaderAdapter(nil, mc, 3*time.Minute, 15*time.Minute)
-		_, err := adapter.UpgradeBackend("missing-node", "llama-cpp", "", "", "", "", 0)
+		_, err := adapter.UpgradeBackend("missing-node", "llama-cpp", "", "", "", "", 0, "", nil)
 		Expect(err).To(HaveOccurred())
 	})
+
+	// Reproducer for "upgrade reports progress:0 the whole time" (Bug B). The
+	// install path streamed per-node download ticks; the upgrade path did a bare
+	// request→single-reply with no progress subscription, so a long force-reinstall
+	// blocked opaque. The adapter must subscribe to the per-op progress subject
+	// (reused from install) BEFORE the request and deliver each tick to onProgress.
+	It("streams per-node progress ticks during the upgrade", func() {
+		mc := newScriptedMessagingClient()
+		nodeID := "node-slow"
+		opID := "op-upgrade-1"
+
+		mc.scriptReply(messaging.SubjectNodeBackendUpgrade(nodeID),
+			messaging.BackendUpgradeReply{Success: true})
+		// The worker would publish these while force-reinstalling. The harness
+		// replays them as soon as the adapter subscribes to the per-op subject.
+		mc.scheduleProgressPublish(nodeID, opID, []messaging.BackendInstallProgressEvent{
+			{NodeID: nodeID, FileName: "llama-cpp.tar", Current: "10 MB", Total: "100 MB", Percentage: 10},
+			{NodeID: nodeID, FileName: "llama-cpp.tar", Current: "100 MB", Total: "100 MB", Percentage: 100},
+		})
+
+		var mu sync.Mutex
+		var got []messaging.BackendInstallProgressEvent
+		onProgress := func(ev messaging.BackendInstallProgressEvent) {
+			mu.Lock()
+			got = append(got, ev)
+			mu.Unlock()
+		}
+
+		adapter := NewRemoteUnloaderAdapter(nil, mc, 3*time.Minute, 15*time.Minute)
+		reply, err := adapter.UpgradeBackend(nodeID, "llama-cpp", `[{"name":"x"}]`, "", "", "", 0, opID, onProgress)
+		Expect(err).ToNot(HaveOccurred())
+		Expect(reply.Success).To(BeTrue())
+
+		// Confirm it subscribed to the (reused) install-progress subject for this op.
+		Expect(mc.subscribeCalls()).To(ContainElement(messaging.SubjectNodeBackendInstallProgress(nodeID, opID)))
+
+		// Progress events are delivered asynchronously (goroutine-per-event), so
+		// poll for both and assert on the set — ordering is best-effort by design.
+		Eventually(func() []float64 {
+			mu.Lock()
+			defer mu.Unlock()
+			pcts := make([]float64, 0, len(got))
+			for _, e := range got {
+				pcts = append(pcts, e.Percentage)
+			}
+			return pcts
+		}, 2*time.Second, 20*time.Millisecond).Should(ConsistOf(float64(10), float64(100)))
+	})
 })