feat(turboquant.cpp): add new backend

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

.agents/adding-backends.md

@@ -0,0 +1,143 @@
+# Adding a New Backend
+
+When adding a new backend to LocalAI, you need to update several files to ensure the backend is properly built, tested, and registered. Here's a step-by-step guide based on the pattern used for adding backends like `moonshine`:
+
+## 1. Create Backend Directory Structure
+
+Create the backend directory under the appropriate location:
+- **Python backends**: `backend/python/<backend-name>/`
+- **Go backends**: `backend/go/<backend-name>/`
+- **C++ backends**: `backend/cpp/<backend-name>/`
+
+For Python backends, you'll typically need:
+- `backend.py` - Main gRPC server implementation
+- `Makefile` - Build configuration
+- `install.sh` - Installation script for dependencies
+- `protogen.sh` - Protocol buffer generation script
+- `requirements.txt` - Python dependencies
+- `run.sh` - Runtime script
+- `test.py` / `test.sh` - Test files
+
+## 2. Add Build Configurations to `.github/workflows/backend.yml`
+
+Add build matrix entries for each platform/GPU type you want to support. Look at similar backends (e.g., `chatterbox`, `faster-whisper`) for reference.
+
+**Placement in file:**
+- CPU builds: Add after other CPU builds (e.g., after `cpu-chatterbox`)
+- CUDA 12 builds: Add after other CUDA 12 builds (e.g., after `gpu-nvidia-cuda-12-chatterbox`)
+- CUDA 13 builds: Add after other CUDA 13 builds (e.g., after `gpu-nvidia-cuda-13-chatterbox`)
+
+**Additional build types you may need:**
+- ROCm/HIP: Use `build-type: 'hipblas'` with `base-image: "rocm/dev-ubuntu-24.04:6.4.4"`
+- Intel/SYCL: Use `build-type: 'intel'` or `build-type: 'sycl_f16'`/`sycl_f32` with `base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"`
+- L4T (ARM): Use `build-type: 'l4t'` with `platforms: 'linux/arm64'` and `runs-on: 'ubuntu-24.04-arm'`
+
+## 3. Add Backend Metadata to `backend/index.yaml`
+
+**Step 3a: Add Meta Definition**
+
+Add a YAML anchor definition in the `## metas` section (around line 2-300). Look for similar backends to use as a template such as `diffusers` or `chatterbox`
+
+**Step 3b: Add Image Entries**
+
+Add image entries at the end of the file, following the pattern of similar backends such as `diffusers` or `chatterbox`. Include both `latest` (production) and `master` (development) tags.
+
+## 4. Update the Makefile
+
+The Makefile needs to be updated in several places to support building and testing the new backend:
+
+**Step 4a: Add to `.NOTPARALLEL`**
+
+Add `backends/<backend-name>` to the `.NOTPARALLEL` line (around line 2) to prevent parallel execution conflicts:
+
+```makefile
+.NOTPARALLEL: ... backends/<backend-name>
+```
+
+**Step 4b: Add to `prepare-test-extra`**
+
+Add the backend to the `prepare-test-extra` target (around line 312) to prepare it for testing:
+
+```makefile
+prepare-test-extra: protogen-python
+	...
+	$(MAKE) -C backend/python/<backend-name>
+```
+
+**Step 4c: Add to `test-extra`**
+
+Add the backend to the `test-extra` target (around line 319) to run its tests:
+
+```makefile
+test-extra: prepare-test-extra
+	...
+	$(MAKE) -C backend/python/<backend-name> test
+```
+
+**Step 4d: Add Backend Definition**
+
+Add a backend definition variable in the backend definitions section (around line 428-457). The format depends on the backend type:
+
+**For Python backends with root context** (like `faster-whisper`, `coqui`):
+```makefile
+BACKEND_<BACKEND_NAME> = <backend-name>|python|.|false|true
+```
+
+**For Python backends with `./backend` context** (like `chatterbox`, `moonshine`):
+```makefile
+BACKEND_<BACKEND_NAME> = <backend-name>|python|./backend|false|true
+```
+
+**For Go backends**:
+```makefile
+BACKEND_<BACKEND_NAME> = <backend-name>|golang|.|false|true
+```
+
+**Step 4e: Generate Docker Build Target**
+
+Add an eval call to generate the docker-build target (around line 480-501):
+
+```makefile
+$(eval $(call generate-docker-build-target,$(BACKEND_<BACKEND_NAME>)))
+```
+
+**Step 4f: Add to `docker-build-backends`**
+
+Add `docker-build-<backend-name>` to the `docker-build-backends` target (around line 507):
+
+```makefile
+docker-build-backends: ... docker-build-<backend-name>
+```
+
+**Determining the Context:**
+
+- If the backend is in `backend/python/<backend-name>/` and uses `./backend` as context in the workflow file, use `./backend` context
+- If the backend is in `backend/python/<backend-name>/` but uses `.` as context in the workflow file, use `.` context
+- Check similar backends to determine the correct context
+
+## 5. Verification Checklist
+
+After adding a new backend, verify:
+
+- [ ] Backend directory structure is complete with all necessary files
+- [ ] Build configurations added to `.github/workflows/backend.yml` for all desired platforms
+- [ ] Meta definition added to `backend/index.yaml` in the `## metas` section
+- [ ] Image entries added to `backend/index.yaml` for all build variants (latest + development)
+- [ ] Tag suffixes match between workflow file and index.yaml
+- [ ] Makefile updated with all 6 required changes (`.NOTPARALLEL`, `prepare-test-extra`, `test-extra`, backend definition, docker-build target eval, `docker-build-backends`)
+- [ ] No YAML syntax errors (check with linter)
+- [ ] No Makefile syntax errors (check with linter)
+- [ ] Follows the same pattern as similar backends (e.g., if it's a transcription backend, follow `faster-whisper` pattern)
+
+## 6. Example: Adding a Python Backend
+
+For reference, when `moonshine` was added:
+- **Files created**: `backend/python/moonshine/{backend.py, Makefile, install.sh, protogen.sh, requirements.txt, run.sh, test.py, test.sh}`
+- **Workflow entries**: 3 build configurations (CPU, CUDA 12, CUDA 13)
+- **Index entries**: 1 meta definition + 6 image entries (cpu, cuda12, cuda13 x latest/development)
+- **Makefile updates**:
+  - Added to `.NOTPARALLEL` line
+  - Added to `prepare-test-extra` and `test-extra` targets
+  - Added `BACKEND_MOONSHINE = moonshine|python|./backend|false|true`
+  - Added eval for docker-build target generation
+  - Added `docker-build-moonshine` to `docker-build-backends`

.agents/api-endpoints-and-auth.md

@@ -0,0 +1,259 @@
+# API Endpoints and Authentication
+
+This guide covers how to add new API endpoints and properly integrate them with the auth/permissions system.
+
+## Architecture overview
+
+Authentication and authorization flow through three layers:
+
+1. **Global auth middleware** (`core/http/auth/middleware.go` → `auth.Middleware`) — applied to every request in `core/http/app.go`. Handles session cookies, Bearer tokens, API keys, and legacy API keys. Populates `auth_user` and `auth_role` in the Echo context.
+2. **Feature middleware** (`auth.RequireFeature`) — per-feature access control applied to route groups or individual routes. Checks if the authenticated user has the specific feature enabled.
+3. **Admin middleware** (`auth.RequireAdmin`) — restricts endpoints to admin users only.
+
+When auth is disabled (no auth DB, no legacy API keys), all middleware becomes pass-through (`auth.NoopMiddleware`).
+
+## Adding a new API endpoint
+
+### Step 1: Create the handler
+
+Write the endpoint handler in the appropriate package under `core/http/endpoints/`. Follow existing patterns:
+
+```go
+// core/http/endpoints/localai/my_feature.go
+func MyFeatureEndpoint(app *application.Application) echo.HandlerFunc {
+    return func(c echo.Context) error {
+        // Use auth.GetUser(c) to get the authenticated user (may be nil if auth is disabled)
+        user := auth.GetUser(c)
+
+        // Your logic here
+        return c.JSON(http.StatusOK, result)
+    }
+}
+```
+
+### Step 2: Register routes
+
+Add routes in the appropriate file under `core/http/routes/`. The file you use depends on the endpoint category:
+
+| File | Category |
+|------|----------|
+| `routes/openai.go` | OpenAI-compatible API endpoints (`/v1/...`) |
+| `routes/localai.go` | LocalAI-specific endpoints (`/api/...`, `/models/...`, `/backends/...`) |
+| `routes/agents.go` | Agent pool endpoints (`/api/agents/...`) |
+| `routes/auth.go` | Auth endpoints (`/api/auth/...`) |
+| `routes/ui_api.go` | UI backend API endpoints |
+
+### Step 3: Apply the right middleware
+
+Choose the appropriate protection level:
+
+#### No auth required (public)
+Exempt paths bypass auth entirely. Add to `isExemptPath()` in `middleware.go` or use the `/api/auth/` prefix (always exempt). Use sparingly — most endpoints should require auth.
+
+#### Standard auth (any authenticated user)
+The global middleware already handles this. API paths (`/api/`, `/v1/`, etc.) automatically require authentication when auth is enabled. You don't need to add any extra middleware.
+
+```go
+router.GET("/v1/my-endpoint", myHandler)  // auth enforced by global middleware
+```
+
+#### Admin only
+Pass `adminMiddleware` to the route. This is set up in `app.go` and passed to `Register*Routes` functions:
+
+```go
+// In the Register function signature, accept the middleware:
+func RegisterMyRoutes(router *echo.Echo, app *application.Application, adminMiddleware echo.MiddlewareFunc) {
+    router.POST("/models/apply", myHandler, adminMiddleware)
+}
+```
+
+#### Feature-gated
+For endpoints that should be toggleable per-user, use feature middleware. There are two approaches:
+
+**Approach A: Route-level middleware** (preferred for groups of related endpoints)
+
+```go
+// In app.go, create the feature middleware:
+myFeatureMw := auth.RequireFeature(application.AuthDB(), auth.FeatureMyFeature)
+
+// Pass it to the route registration function:
+routes.RegisterMyRoutes(e, app, myFeatureMw)
+
+// In the routes file, apply to a group:
+g := e.Group("/api/my-feature", myFeatureMw)
+g.GET("", listHandler)
+g.POST("", createHandler)
+```
+
+**Approach B: RouteFeatureRegistry** (preferred for individual OpenAI-compatible endpoints)
+
+Add an entry to `RouteFeatureRegistry` in `core/http/auth/features.go`. The `RequireRouteFeature` global middleware will automatically enforce it:
+
+```go
+var RouteFeatureRegistry = []RouteFeature{
+    // ... existing entries ...
+    {"POST", "/v1/my-endpoint", FeatureMyFeature},
+}
+```
+
+## Adding a new feature
+
+When you need a new toggleable feature (not just a new endpoint under an existing feature):
+
+### 1. Define the feature constant
+
+Add to `core/http/auth/permissions.go`:
+
+```go
+const (
+    // Add to the appropriate group:
+    // Agent features (default OFF for new users)
+    FeatureMyFeature = "my_feature"
+
+    // OR API features (default ON for new users)
+    FeatureMyFeature = "my_feature"
+)
+```
+
+Then add it to the appropriate slice:
+
+```go
+// Default OFF — user must be explicitly granted access:
+var AgentFeatures = []string{..., FeatureMyFeature}
+
+// Default ON — user has access unless explicitly revoked:
+var APIFeatures = []string{..., FeatureMyFeature}
+```
+
+### 2. Add feature metadata
+
+In `core/http/auth/features.go`, add to the appropriate `FeatureMetas` function so the admin UI can display it:
+
+```go
+func AgentFeatureMetas() []FeatureMeta {
+    return []FeatureMeta{
+        // ... existing ...
+        {FeatureMyFeature, "My Feature", false},  // false = default OFF
+    }
+}
+```
+
+### 3. Wire up the middleware
+
+In `core/http/app.go`:
+
+```go
+myFeatureMw := auth.RequireFeature(application.AuthDB(), auth.FeatureMyFeature)
+```
+
+Then pass it to the route registration function.
+
+### 4. Register route-feature mappings (if applicable)
+
+If your feature gates standard API endpoints (like `/v1/...`), add entries to `RouteFeatureRegistry` in `features.go` instead of using per-route middleware.
+
+## Accessing the authenticated user in handlers
+
+```go
+import "github.com/mudler/LocalAI/core/http/auth"
+
+func MyHandler(c echo.Context) error {
+    // Get the user (nil when auth is disabled or unauthenticated)
+    user := auth.GetUser(c)
+    if user == nil {
+        // Handle unauthenticated — or let middleware handle it
+    }
+
+    // Check role
+    if user.Role == auth.RoleAdmin {
+        // admin-specific logic
+    }
+
+    // Check feature access programmatically (when you need conditional behavior, not full blocking)
+    if auth.HasFeatureAccess(db, user, auth.FeatureMyFeature) {
+        // feature-specific logic
+    }
+
+    // Check model access
+    if !auth.IsModelAllowed(db, user, modelName) {
+        return c.JSON(http.StatusForbidden, ...)
+    }
+}
+```
+
+## Middleware composition patterns
+
+Middleware can be composed at different levels. Here are the patterns used in the codebase:
+
+### Group-level middleware (agents pattern)
+```go
+// All routes in the group share the middleware
+g := e.Group("/api/agents", poolReadyMw, agentsMw)
+g.GET("", listHandler)
+g.POST("", createHandler)
+```
+
+### Per-route middleware (localai pattern)
+```go
+// Individual routes get middleware as extra arguments
+router.POST("/models/apply", applyHandler, adminMiddleware)
+router.GET("/metrics", metricsHandler, adminMiddleware)
+```
+
+### Middleware slice (openai pattern)
+```go
+// Build a middleware chain for a handler
+chatMiddleware := []echo.MiddlewareFunc{
+    usageMiddleware,
+    traceMiddleware,
+    modelFilterMiddleware,
+}
+app.POST("/v1/chat/completions", chatHandler, chatMiddleware...)
+```
+
+## Error response format
+
+Always use `schema.ErrorResponse` for auth/permission errors to stay consistent with the OpenAI-compatible API:
+
+```go
+return c.JSON(http.StatusForbidden, schema.ErrorResponse{
+    Error: &schema.APIError{
+        Message: "feature not enabled for your account",
+        Code:    http.StatusForbidden,
+        Type:    "authorization_error",
+    },
+})
+```
+
+Use these HTTP status codes:
+- `401 Unauthorized` — no valid credentials provided
+- `403 Forbidden` — authenticated but lacking permission
+- `429 Too Many Requests` — rate limited (auth endpoints)
+
+## Usage tracking
+
+If your endpoint should be tracked for usage (token counts, request counts), add the `usageMiddleware` to its middleware chain. See `core/http/middleware/usage.go` and how it's applied in `routes/openai.go`.
+
+## Path protection rules
+
+The global auth middleware classifies paths as API paths or non-API paths:
+
+- **API paths** (always require auth when auth is enabled): `/api/`, `/v1/`, `/models/`, `/backends/`, `/backend/`, `/tts`, `/vad`, `/video`, `/stores/`, `/system`, `/ws/`, `/metrics`
+- **Exempt paths** (never require auth): `/api/auth/` prefix, anything in `appConfig.PathWithoutAuth`
+- **Non-API paths** (UI, static assets): pass through without auth — the React UI handles login redirects client-side
+
+If you add endpoints under a new top-level path prefix, add it to `isAPIPath()` in `middleware.go` to ensure it requires authentication.
+
+## Checklist
+
+When adding a new endpoint:
+
+- [ ] Handler in `core/http/endpoints/`
+- [ ] Route registered in appropriate `core/http/routes/` file
+- [ ] Auth level chosen: public / standard / admin / feature-gated
+- [ ] If feature-gated: constant in `permissions.go`, metadata in `features.go`, middleware in `app.go`
+- [ ] If new path prefix: added to `isAPIPath()` in `middleware.go`
+- [ ] If OpenAI-compatible: entry in `RouteFeatureRegistry`
+- [ ] If token-counting: `usageMiddleware` added to middleware chain
+- [ ] Error responses use `schema.ErrorResponse` format
+- [ ] Tests cover both authenticated and unauthenticated access

.agents/building-and-testing.md

@@ -0,0 +1,16 @@
+# Build and Testing
+
+Building and testing the project depends on the components involved and the platform where development is taking place. Due to the amount of context required it's usually best not to try building or testing the project unless the user requests it. If you must build the project then inspect the Makefile in the project root and the Makefiles of any backends that are effected by changes you are making. In addition the workflows in .github/workflows can be used as a reference when it is unclear how to build or test a component. The primary Makefile contains targets for building inside or outside Docker, if the user has not previously specified a preference then ask which they would like to use.
+
+## Building a specified backend
+
+Let's say the user wants to build a particular backend for a given platform. For example let's say they want to build coqui for ROCM/hipblas
+
+- The Makefile has targets like `docker-build-coqui` created with `generate-docker-build-target` at the time of writing. Recently added backends may require a new target.
+- At a minimum we need to set the BUILD_TYPE, BASE_IMAGE build-args
+  - Use .github/workflows/backend.yml as a reference it lists the needed args in the `include` job strategy matrix
+  - l4t and cublas also requires the CUDA major and minor version
+- You can pretty print a command like `DOCKER_MAKEFLAGS=-j$(nproc --ignore=1) BUILD_TYPE=hipblas BASE_IMAGE=rocm/dev-ubuntu-24.04:6.4.4 make docker-build-coqui`
+- Unless the user specifies that they want you to run the command, then just print it because not all agent frontends handle long running jobs well and the output may overflow your context
+- The user may say they want to build AMD or ROCM instead of hipblas, or Intel instead of SYCL or NVIDIA insted of l4t or cublas. Ask for confirmation if there is ambiguity.
+- Sometimes the user may need extra parameters to be added to `docker build` (e.g. `--platform` for cross-platform builds or `--progress` to view the full logs), in which case you can generate the `docker build` command directly.

.agents/coding-style.md

@@ -0,0 +1,52 @@
+# Coding Style
+
+The project has the following .editorconfig:
+
+```
+root = true
+
+[*]
+indent_style = space
+indent_size = 2
+end_of_line = lf
+charset = utf-8
+trim_trailing_whitespace = true
+insert_final_newline = true
+
+[*.go]
+indent_style = tab
+
+[Makefile]
+indent_style = tab
+
+[*.proto]
+indent_size = 2
+
+[*.py]
+indent_size = 4
+
+[*.js]
+indent_size = 2
+
+[*.yaml]
+indent_size = 2
+
+[*.md]
+trim_trailing_whitespace = false
+```
+
+- Use comments sparingly to explain why code does something, not what it does. Comments are there to add context that would be difficult to deduce from reading the code.
+- Prefer modern Go e.g. use `any` not `interface{}`
+
+## Logging
+
+Use `github.com/mudler/xlog` for logging which has the same API as slog.
+
+## Documentation
+
+The project documentation is located in `docs/content`. When adding new features or changing existing functionality, it is crucial to update the documentation to reflect these changes. This helps users understand how to use the new capabilities and ensures the documentation stays relevant.
+
+- **Feature Documentation**: If you add a new feature (like a new backend or API endpoint), create a new markdown file in `docs/content/features/` explaining what it is, how to configure it, and how to use it.
+- **Configuration**: If you modify configuration options, update the relevant sections in `docs/content/`.
+- **Examples**: providing concrete examples (like YAML configuration blocks) is highly encouraged to help users get started quickly.
+- **Shortcodes**: Use `{{% notice note %}}`, `{{% notice tip %}}`, or `{{% notice warning %}}` for callout boxes. Do **not** use `{{% alert %}}` — that shortcode does not exist in this project's Hugo theme and will break the docs build.

.agents/debugging-backends.md

@@ -0,0 +1,141 @@
+# Debugging and Rebuilding Backends
+
+When a backend fails at runtime (e.g. a gRPC method error, a Python import error, or a dependency conflict), use this guide to diagnose, fix, and rebuild.
+
+## Architecture Overview
+
+- **Source directory**: `backend/python/<name>/` (or `backend/go/<name>/`, `backend/cpp/<name>/`)
+- **Installed directory**: `backends/<name>/` — this is what LocalAI actually runs. It is populated by `make backends/<name>` which builds a Docker image, exports it, and installs it via `local-ai backends install`.
+- **Virtual environment**: `backends/<name>/venv/` — the installed Python venv (for Python backends). The Python binary is at `backends/<name>/venv/bin/python`.
+
+Editing files in `backend/python/<name>/` does **not** affect the running backend until you rebuild with `make backends/<name>`.
+
+## Diagnosing Failures
+
+### 1. Check the logs
+
+Backend gRPC processes log to LocalAI's stdout/stderr. Look for lines tagged with the backend's model ID:
+
+```
+GRPC stderr id="trl-finetune-127.0.0.1:37335" line="..."
+```
+
+Common error patterns:
+- **"Method not implemented"** — the backend is missing a gRPC method that the Go side calls. The model loader (`pkg/model/initializers.go`) always calls `LoadModel` after `Health`; fine-tuning backends must implement it even as a no-op stub.
+- **Python import errors / `AttributeError`** — usually a dependency version mismatch (e.g. `pyarrow` removing `PyExtensionType`).
+- **"failed to load backend"** — the gRPC process crashed or never started. Check stderr lines for the traceback.
+
+### 2. Test the Python environment directly
+
+You can run the installed venv's Python to check imports without starting the full server:
+
+```bash
+backends/<name>/venv/bin/python -c "import datasets; print(datasets.__version__)"
+```
+
+If `pip` is missing from the venv, bootstrap it:
+
+```bash
+backends/<name>/venv/bin/python -m ensurepip
+```
+
+Then use `backends/<name>/venv/bin/python -m pip install ...` to test fixes in the installed venv before committing them to the source requirements.
+
+### 3. Check upstream dependency constraints
+
+When you hit a dependency conflict, check what the main library expects. For example, TRL's upstream `requirements.txt`:
+
+```
+https://github.com/huggingface/trl/blob/main/requirements.txt
+```
+
+Pin minimum versions in the backend's requirements files to match upstream.
+
+## Common Fixes
+
+### Missing gRPC methods
+
+If the Go side calls a method the backend doesn't implement (e.g. `LoadModel`), add a no-op stub in `backend.py`:
+
+```python
+def LoadModel(self, request, context):
+    """No-op — actual loading happens elsewhere."""
+    return backend_pb2.Result(success=True, message="OK")
+```
+
+The gRPC contract requires `LoadModel` to succeed for the model loader to return a usable client, even if the backend doesn't need upfront model loading.
+
+### Dependency version conflicts
+
+Python backends often break when a transitive dependency releases a breaking change (e.g. `pyarrow` removing `PyExtensionType`). Steps:
+
+1. Identify the broken import in the logs
+2. Test in the installed venv: `backends/<name>/venv/bin/python -c "import <module>"`
+3. Check upstream requirements for version constraints
+4. Update **all** requirements files in `backend/python/<name>/`:
+   - `requirements.txt` — base deps (grpcio, protobuf)
+   - `requirements-cpu.txt` — CPU-specific (includes PyTorch CPU index)
+   - `requirements-cublas12.txt` — CUDA 12
+   - `requirements-cublas13.txt` — CUDA 13
+5. Rebuild: `make backends/<name>`
+
+### PyTorch index conflicts (uv resolver)
+
+The Docker build uses `uv` for pip installs. When `--extra-index-url` points to the PyTorch wheel index, `uv` may refuse to fetch packages like `requests` from PyPI if it finds a different version on the PyTorch index first. Fix this by adding `--index-strategy=unsafe-first-match` to `install.sh`:
+
+```bash
+EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match"
+installRequirements
+```
+
+Most Python backends already do this — check `backend/python/transformers/install.sh` or similar for reference.
+
+## Rebuilding
+
+### Rebuild a single backend
+
+```bash
+make backends/<name>
+```
+
+This runs the Docker build (`Dockerfile.python`), exports the image to `backend-images/<name>.tar`, and installs it into `backends/<name>/`. It also rebuilds the `local-ai` Go binary (without extra tags).
+
+**Important**: If you were previously running with `GO_TAGS=auth`, the `make backends/<name>` step will overwrite your binary without that tag. Rebuild the Go binary afterward:
+
+```bash
+GO_TAGS=auth make build
+```
+
+### Rebuild and restart
+
+After rebuilding a backend, you must restart LocalAI for it to pick up the new backend files. The backend gRPC process is spawned on demand when the model is first loaded.
+
+```bash
+# Kill existing process
+kill <pid>
+
+# Restart
+./local-ai run --debug [your flags]
+```
+
+### Quick iteration (skip Docker rebuild)
+
+For fast iteration on a Python backend's `backend.py` without a full Docker rebuild, you can edit the installed copy directly:
+
+```bash
+# Edit the installed copy
+vim backends/<name>/backend.py
+
+# Restart LocalAI to respawn the gRPC process
+```
+
+This is useful for testing but **does not persist** — the next `make backends/<name>` will overwrite it. Always commit fixes to the source in `backend/python/<name>/`.
+
+## Verification
+
+After fixing and rebuilding:
+
+1. Start LocalAI and confirm the backend registers: look for `Registering backend name="<name>"` in the logs
+2. Trigger the operation that failed (e.g. start a fine-tuning job)
+3. Watch the GRPC stderr/stdout lines for the backend's model ID
+4. Confirm no errors in the traceback

.agents/llama-cpp-backend.md

@@ -0,0 +1,77 @@
+# llama.cpp Backend
+
+The llama.cpp backend (`backend/cpp/llama-cpp/grpc-server.cpp`) is a gRPC adaptation of the upstream HTTP server (`llama.cpp/tools/server/server.cpp`). It uses the same underlying server infrastructure from `llama.cpp/tools/server/server-context.cpp`.
+
+## Building and Testing
+
+- Test llama.cpp backend compilation: `make backends/llama-cpp`
+- The backend is built as part of the main build process
+- Check `backend/cpp/llama-cpp/Makefile` for build configuration
+
+## Architecture
+
+- **grpc-server.cpp**: gRPC server implementation, adapts HTTP server patterns to gRPC
+- Uses shared server infrastructure: `server-context.cpp`, `server-task.cpp`, `server-queue.cpp`, `server-common.cpp`
+- The gRPC server mirrors the HTTP server's functionality but uses gRPC instead of HTTP
+
+## Common Issues When Updating llama.cpp
+
+When fixing compilation errors after upstream changes:
+1. Check how `server.cpp` (HTTP server) handles the same change
+2. Look for new public APIs or getter methods
+3. Store copies of needed data instead of accessing private members
+4. Update function calls to match new signatures
+5. Test with `make backends/llama-cpp`
+
+## Key Differences from HTTP Server
+
+- gRPC uses `BackendServiceImpl` class with gRPC service methods
+- HTTP server uses `server_routes` with HTTP handlers
+- Both use the same `server_context` and task queue infrastructure
+- gRPC methods: `LoadModel`, `Predict`, `PredictStream`, `Embedding`, `Rerank`, `TokenizeString`, `GetMetrics`, `Health`
+
+## Tool Call Parsing Maintenance
+
+When working on JSON/XML tool call parsing functionality, always check llama.cpp for reference implementation and updates:
+
+### Checking for XML Parsing Changes
+
+1. **Review XML Format Definitions**: Check `llama.cpp/common/chat-parser-xml-toolcall.h` for `xml_tool_call_format` struct changes
+2. **Review Parsing Logic**: Check `llama.cpp/common/chat-parser-xml-toolcall.cpp` for parsing algorithm updates
+3. **Review Format Presets**: Check `llama.cpp/common/chat-parser.cpp` for new XML format presets (search for `xml_tool_call_format form`)
+4. **Review Model Lists**: Check `llama.cpp/common/chat.h` for `COMMON_CHAT_FORMAT_*` enum values that use XML parsing:
+   - `COMMON_CHAT_FORMAT_GLM_4_5`
+   - `COMMON_CHAT_FORMAT_MINIMAX_M2`
+   - `COMMON_CHAT_FORMAT_KIMI_K2`
+   - `COMMON_CHAT_FORMAT_QWEN3_CODER_XML`
+   - `COMMON_CHAT_FORMAT_APRIEL_1_5`
+   - `COMMON_CHAT_FORMAT_XIAOMI_MIMO`
+   - Any new formats added
+
+### Model Configuration Options
+
+Always check `llama.cpp` for new model configuration options that should be supported in LocalAI:
+
+1. **Check Server Context**: Review `llama.cpp/tools/server/server-context.cpp` for new parameters
+2. **Check Chat Params**: Review `llama.cpp/common/chat.h` for `common_chat_params` struct changes
+3. **Check Server Options**: Review `llama.cpp/tools/server/server.cpp` for command-line argument changes
+4. **Examples of options to check**:
+   - `ctx_shift` - Context shifting support
+   - `parallel_tool_calls` - Parallel tool calling
+   - `reasoning_format` - Reasoning format options
+   - Any new flags or parameters
+
+### Implementation Guidelines
+
+1. **Feature Parity**: Always aim for feature parity with llama.cpp's implementation
+2. **Test Coverage**: Add tests for new features matching llama.cpp's behavior
+3. **Documentation**: Update relevant documentation when adding new formats or options
+4. **Backward Compatibility**: Ensure changes don't break existing functionality
+
+### Files to Monitor
+
+- `llama.cpp/common/chat-parser-xml-toolcall.h` - Format definitions
+- `llama.cpp/common/chat-parser-xml-toolcall.cpp` - Parsing logic
+- `llama.cpp/common/chat-parser.cpp` - Format presets and model-specific handlers
+- `llama.cpp/common/chat.h` - Format enums and parameter structures
+- `llama.cpp/tools/server/server-context.cpp` - Server configuration options

.agents/testing-mcp-apps.md

@@ -0,0 +1,120 @@
+# Testing MCP Apps (Interactive Tool UIs)
+
+MCP Apps is an extension to MCP where tools declare interactive HTML UIs via `_meta.ui.resourceUri`. When the LLM calls such a tool, the UI renders the app in a sandboxed iframe inline in the chat. The app communicates bidirectionally with the host via `postMessage` (JSON-RPC) and can call server tools, send messages, and update model context.
+
+Spec: https://modelcontextprotocol.io/extensions/apps/overview
+
+## Quick Start: Run a Test MCP App Server
+
+The `@modelcontextprotocol/server-basic-react` npm package is a ready-to-use test server that exposes a `get-time` tool with an interactive React clock UI. It requires Node >= 20, so run it in Docker:
+
+```bash
+docker run -d --name mcp-app-test -p 3001:3001 node:22-slim \
+  sh -c 'npx -y @modelcontextprotocol/server-basic-react'
+```
+
+Wait ~10 seconds for it to start, then verify:
+
+```bash
+# Check it's running
+docker logs mcp-app-test
+# Expected: "MCP server listening on http://localhost:3001/mcp"
+
+# Verify MCP protocol works
+curl -s -X POST http://localhost:3001/mcp \
+  -H 'Content-Type: application/json' \
+  -H 'Accept: application/json, text/event-stream' \
+  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-03-26","capabilities":{},"clientInfo":{"name":"test","version":"1.0.0"}}}'
+
+# List tools — should show get-time with _meta.ui.resourceUri
+curl -s -X POST http://localhost:3001/mcp \
+  -H 'Content-Type: application/json' \
+  -H 'Accept: application/json, text/event-stream' \
+  -d '{"jsonrpc":"2.0","id":2,"method":"tools/list","params":{}}'
+```
+
+The `tools/list` response should contain:
+```json
+{
+  "name": "get-time",
+  "_meta": {
+    "ui": { "resourceUri": "ui://get-time/mcp-app.html" }
+  }
+}
+```
+
+## Testing in LocalAI's UI
+
+1. Make sure LocalAI is running (e.g. `http://localhost:8080`)
+2. Build the React UI: `cd core/http/react-ui && npm install && npm run build`
+3. Open the Chat page in your browser
+4. Click **"Client MCP"** in the chat header
+5. Add a new client MCP server:
+   - **URL**: `http://localhost:3001/mcp`
+   - **Use CORS proxy**: enabled (default) — required because the browser can't hit `localhost:3001` directly due to CORS; LocalAI's proxy at `/api/cors-proxy` handles it
+6. The server should connect and discover the `get-time` tool
+7. Select a model and send: **"What time is it?"**
+8. The LLM should call the `get-time` tool
+9. The tool result should render the interactive React clock app in an iframe as a standalone chat message (not inside the collapsed activity group)
+
+## What to Verify
+
+- [ ] Tool appears in the connected tools list (not filtered — `get-time` is callable by the LLM)
+- [ ] The iframe renders as a standalone chat message with a puzzle-piece icon
+- [ ] The app loads and is interactive (clock UI, buttons work)
+- [ ] No "Reconnect to MCP server" overlay (connection is live)
+- [ ] Console logs show bidirectional communication:
+  - `tools/call` messages from app to host (app calling server tools)
+  - `ui/message` notifications (app sending messages)
+- [ ] After the app renders, the LLM continues and produces a text response with the time
+- [ ] Non-UI tools continue to work normally (text-only results)
+- [ ] Page reload shows the HTML statically with a reconnect overlay until you reconnect
+
+## Console Log Patterns
+
+Healthy bidirectional communication looks like:
+
+```
+Parsed message { jsonrpc: "2.0", id: N, result: {...} }     // Bridge init
+get-time result: { content: [...] }                          // Tool result received
+Calling get-time tool...                                     // App calls tool
+Sending message { method: "tools/call", ... }                // App -> host -> server
+Parsed message { jsonrpc: "2.0", id: N, result: {...} }     // Server response
+Sending message text to Host: ...                            // App sends message
+Sending message { method: "ui/message", ... }                // Message notification
+Message accepted                                             // Host acknowledged
+```
+
+Benign warnings to ignore:
+- `Source map error: ... about:srcdoc` — browser devtools can't find source maps for srcdoc iframes
+- `Ignoring message from unknown source` — duplicate postMessage from iframe navigation
+- `notifications/cancelled` — app cleaning up previous requests
+
+## Architecture Notes
+
+- **No server-side changes needed** — the MCP App protocol runs entirely in the browser
+- `PostMessageTransport` wraps `window.postMessage` between host and `srcdoc` iframe
+- `AppBridge` (from `@modelcontextprotocol/ext-apps`) auto-forwards `tools/call`, `resources/read`, `resources/list` from the app to the MCP server via the host's `Client`
+- The iframe uses `sandbox="allow-scripts allow-forms"` (no `allow-same-origin`) — opaque origin, no access to host cookies/DOM/localStorage
+- App-only tools (`_meta.ui.visibility: "app-only"`) are filtered from the LLM's tool list but remain callable by the app iframe
+
+## Key Files
+
+- `core/http/react-ui/src/components/MCPAppFrame.jsx` — iframe + AppBridge component
+- `core/http/react-ui/src/hooks/useMCPClient.js` — MCP client hook with app UI helpers (`hasAppUI`, `getAppResource`, `getClientForTool`, `getToolDefinition`)
+- `core/http/react-ui/src/hooks/useChat.js` — agentic loop, attaches `appUI` to tool_result messages
+- `core/http/react-ui/src/pages/Chat.jsx` — renders MCPAppFrame as standalone chat messages
+
+## Other Test Servers
+
+The `@modelcontextprotocol/ext-apps` repo has many example servers:
+- `@modelcontextprotocol/server-basic-react` — simple clock (React)
+- More examples at https://github.com/modelcontextprotocol/ext-apps/tree/main/examples
+
+All examples support both stdio and HTTP transport. Run without `--stdio` for HTTP mode on port 3001.
+
+## Cleanup
+
+```bash
+docker rm -f mcp-app-test
+```

.devcontainer/docker-compose-devcontainer.yml

@@ -10,7 +10,8 @@ services:
       - 8080:8080
     volumes:
       - localai_workspace:/workspace
-      - ../models:/host-models
+      - models:/host-models
+      - backends:/host-backends
       - ./customization:/devcontainer-customization
     command: /bin/sh -c "while sleep 1000; do :; done"
     cap_add:
@@ -39,6 +40,9 @@ services:
       - GF_SECURITY_ADMIN_PASSWORD=grafana
     volumes:
       - ./grafana:/etc/grafana/provisioning/datasources
+
 volumes:
   prom_data:
-  localai_workspace:
+  localai_workspace:
+  models:
+  backends:

.env

@@ -26,6 +26,9 @@
 ## Disables COMPEL (Diffusers)
 # COMPEL=0
 
+## Disables SD_EMBED (Diffusers)
+# SD_EMBED=0
+
 ## Enable/Disable single backend (useful if only one GPU is available)
 # LOCALAI_SINGLE_ACTIVE_BACKEND=true
 

.github/gallery-agent/agent.go

@@ -13,8 +13,8 @@ import (
 
 	"github.com/ghodss/yaml"
 	hfapi "github.com/mudler/LocalAI/pkg/huggingface-api"
-	cogito "github.com/mudler/cogito"
-
+	"github.com/mudler/cogito"
+	"github.com/mudler/cogito/clients"
 	"github.com/mudler/cogito/structures"
 	"github.com/sashabaranov/go-openai/jsonschema"
 )
@@ -25,7 +25,7 @@ var (
 	openAIBaseURL    = os.Getenv("OPENAI_BASE_URL")
 	galleryIndexPath = os.Getenv("GALLERY_INDEX_PATH")
 	//defaultclient
-	llm = cogito.NewOpenAILLM(openAIModel, openAIKey, openAIBaseURL)
+	llm = clients.NewOpenAILLM(openAIModel, openAIKey, openAIBaseURL)
 )
 
 // cleanTextContent removes trailing spaces, tabs, and normalizes line endings
@@ -133,6 +133,7 @@ func getRealReadme(ctx context.Context, repository string) (string, error) {
 	result, err := cogito.ExecuteTools(llm, fragment,
 		cogito.WithIterations(3),
 		cogito.WithMaxAttempts(3),
+		cogito.DisableSinkState,
 		cogito.WithTools(&HFReadmeTool{client: hfapi.NewClient()}))
 	if err != nil {
 		return "", err
@@ -141,12 +142,12 @@ func getRealReadme(ctx context.Context, repository string) (string, error) {
 	result = result.AddMessage("user", "Describe the model in a clear and concise way that can be shared in a model gallery.")
 
 	// Get a response
-	newFragment, err := llm.Ask(ctx, result)
+	_, err = llm.Ask(ctx, result)
 	if err != nil {
 		return "", err
 	}
 
-	content := newFragment.LastMessage().Content
+	content := result.LastMessage().Content
 	return cleanTextContent(content), nil
 }
 
@@ -406,7 +407,7 @@ func getHuggingFaceAvatarURL(author string) string {
 	}
 
 	// Parse the response to get avatar URL
-	var userInfo map[string]interface{}
+	var userInfo map[string]any
 	body, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return ""

.github/gallery-agent/gallery.go

@@ -79,7 +79,20 @@ func generateYAMLEntry(model ProcessedModel, quantization string) string {
 	description = cleanTextContent(description)
 	formattedDescription := formatTextContent(description)
 
-	configFile := formatTextContent(modelConfig.ConfigFile)
+	// Strip name and description from config file since they are
+	// already present at the gallery entry level and should not
+	// appear under overrides.
+	configFileContent := modelConfig.ConfigFile
+	var cfgMap map[string]any
+	if err := yaml.Unmarshal([]byte(configFileContent), &cfgMap); err == nil {
+		delete(cfgMap, "name")
+		delete(cfgMap, "description")
+		if cleaned, err := yaml.Marshal(cfgMap); err == nil {
+			configFileContent = string(cleaned)
+		}
+	}
+
+	configFile := formatTextContent(configFileContent)
 
 	filesYAML, _ := yaml.Marshal(modelConfig.Files)
 

.github/gallery-agent/testing.go

@@ -3,7 +3,7 @@ package main
 import (
 	"context"
 	"fmt"
-	"math/rand"
+	"math/rand/v2"
 	"strings"
 	"time"
 )
@@ -13,11 +13,11 @@ func runSyntheticMode() error {
 	generator := NewSyntheticDataGenerator()
 
 	// Generate a random number of synthetic models (1-3)
-	numModels := generator.rand.Intn(3) + 1
+	numModels := generator.rand.IntN(3) + 1
 	fmt.Printf("Generating %d synthetic models for testing...\n", numModels)
 
 	var models []ProcessedModel
-	for i := 0; i < numModels; i++ {
+	for range numModels {
 		model := generator.GenerateProcessedModel()
 		models = append(models, model)
 		fmt.Printf("Generated synthetic model: %s\n", model.ModelID)
@@ -42,14 +42,14 @@ type SyntheticDataGenerator struct {
 // NewSyntheticDataGenerator creates a new synthetic data generator
 func NewSyntheticDataGenerator() *SyntheticDataGenerator {
 	return &SyntheticDataGenerator{
-		rand: rand.New(rand.NewSource(time.Now().UnixNano())),
+		rand: rand.New(rand.NewPCG(uint64(time.Now().UnixNano()), 0)),
 	}
 }
 
 // GenerateProcessedModelFile creates a synthetic ProcessedModelFile
 func (g *SyntheticDataGenerator) GenerateProcessedModelFile() ProcessedModelFile {
 	fileTypes := []string{"model", "readme", "other"}
-	fileType := fileTypes[g.rand.Intn(len(fileTypes))]
+	fileType := fileTypes[g.rand.IntN(len(fileTypes))]
 
 	var path string
 	var isReadme bool
@@ -68,7 +68,7 @@ func (g *SyntheticDataGenerator) GenerateProcessedModelFile() ProcessedModelFile
 
 	return ProcessedModelFile{
 		Path:     path,
-		Size:     int64(g.rand.Intn(1000000000) + 1000000), // 1MB to 1GB
+		Size:     int64(g.rand.IntN(1000000000) + 1000000), // 1MB to 1GB
 		SHA256:   g.randomSHA256(),
 		IsReadme: isReadme,
 		FileType: fileType,
@@ -80,19 +80,19 @@ func (g *SyntheticDataGenerator) GenerateProcessedModel() ProcessedModel {
 	authors := []string{"microsoft", "meta", "google", "openai", "anthropic", "mistralai", "huggingface"}
 	modelNames := []string{"llama", "gpt", "claude", "mistral", "gemma", "phi", "qwen", "codellama"}
 
-	author := authors[g.rand.Intn(len(authors))]
-	modelName := modelNames[g.rand.Intn(len(modelNames))]
+	author := authors[g.rand.IntN(len(authors))]
+	modelName := modelNames[g.rand.IntN(len(modelNames))]
 	modelID := fmt.Sprintf("%s/%s-%s", author, modelName, g.randomString(6))
 
 	// Generate files
-	numFiles := g.rand.Intn(5) + 2 // 2-6 files
+	numFiles := g.rand.IntN(5) + 2 // 2-6 files
 	files := make([]ProcessedModelFile, numFiles)
 
 	// Ensure at least one model file and one readme
 	hasModelFile := false
 	hasReadme := false
 
-	for i := 0; i < numFiles; i++ {
+	for i := range numFiles {
 		files[i] = g.GenerateProcessedModelFile()
 		if files[i].FileType == "model" {
 			hasModelFile = true
@@ -140,27 +140,27 @@ func (g *SyntheticDataGenerator) GenerateProcessedModel() ProcessedModel {
 
 	// Generate sample metadata
 	licenses := []string{"apache-2.0", "mit", "llama2", "gpl-3.0", "bsd", ""}
-	license := licenses[g.rand.Intn(len(licenses))]
+	license := licenses[g.rand.IntN(len(licenses))]
 
 	sampleTags := []string{"llm", "gguf", "gpu", "cpu", "text-to-text", "chat", "instruction-tuned"}
-	numTags := g.rand.Intn(4) + 3 // 3-6 tags
+	numTags := g.rand.IntN(4) + 3 // 3-6 tags
 	tags := make([]string, numTags)
-	for i := 0; i < numTags; i++ {
-		tags[i] = sampleTags[g.rand.Intn(len(sampleTags))]
+	for i := range numTags {
+		tags[i] = sampleTags[g.rand.IntN(len(sampleTags))]
 	}
 	// Remove duplicates
 	tags = g.removeDuplicates(tags)
 
 	// Optionally include icon (50% chance)
 	icon := ""
-	if g.rand.Intn(2) == 0 {
+	if g.rand.IntN(2) == 0 {
 		icon = fmt.Sprintf("https://cdn-avatars.huggingface.co/v1/production/uploads/%s.png", g.randomString(24))
 	}
 
 	return ProcessedModel{
 		ModelID:                 modelID,
 		Author:                  author,
-		Downloads:               g.rand.Intn(1000000) + 1000,
+		Downloads:               g.rand.IntN(1000000) + 1000,
 		LastModified:            g.randomDate(),
 		Files:                   files,
 		PreferredModelFile:      preferredModelFile,
@@ -180,7 +180,7 @@ func (g *SyntheticDataGenerator) randomString(length int) string {
 	const charset = "abcdefghijklmnopqrstuvwxyz0123456789"
 	b := make([]byte, length)
 	for i := range b {
-		b[i] = charset[g.rand.Intn(len(charset))]
+		b[i] = charset[g.rand.IntN(len(charset))]
 	}
 	return string(b)
 }
@@ -189,14 +189,14 @@ func (g *SyntheticDataGenerator) randomSHA256() string {
 	const charset = "0123456789abcdef"
 	b := make([]byte, 64)
 	for i := range b {
-		b[i] = charset[g.rand.Intn(len(charset))]
+		b[i] = charset[g.rand.IntN(len(charset))]
 	}
 	return string(b)
 }
 
 func (g *SyntheticDataGenerator) randomDate() string {
 	now := time.Now()
-	daysAgo := g.rand.Intn(365) // Random date within last year
+	daysAgo := g.rand.IntN(365) // Random date within last year
 	pastDate := now.AddDate(0, 0, -daysAgo)
 	return pastDate.Format("2006-01-02T15:04:05.000Z")
 }
@@ -220,5 +220,5 @@ func (g *SyntheticDataGenerator) generateReadmeContent(modelName, author string)
 		fmt.Sprintf("# %s Language Model\n\nDeveloped by %s, this model represents state-of-the-art performance in natural language understanding and generation.\n\n## Key Features\n\n- Multilingual support\n- Context-aware responses\n- Efficient memory usage\n- Fast inference speed\n\n## Applications\n\n- Chatbots and virtual assistants\n- Content generation\n- Code completion\n- Educational tools", strings.Title(modelName), author),
 	}
 
-	return templates[g.rand.Intn(len(templates))]
+	return templates[g.rand.IntN(len(templates))]
 }

.github/gallery-agent/tools.go

@@ -13,16 +13,16 @@ type HFReadmeTool struct {
 	client *hfapi.Client
 }
 
-func (s *HFReadmeTool) Execute(args map[string]any) (string, error) {
+func (s *HFReadmeTool) Execute(args map[string]any) (string, any, error) {
 	q, ok := args["repository"].(string)
 	if !ok {
-		return "", fmt.Errorf("no query")
+		return "", nil, fmt.Errorf("no query")
 	}
 	readme, err := s.client.GetReadmeContent(q, "README.md")
 	if err != nil {
-		return "", err
+		return "", nil, err
 	}
-	return readme, nil
+	return readme, nil, nil
 }
 
 func (s *HFReadmeTool) Tool() openai.Tool {

.github/workflows/backend.yml

@@ -118,6 +118,32 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-trl'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "trl"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-llama-cpp-quantization'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "llama-cpp-quantization"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -157,6 +183,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-mlx-distributed'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "mlx-distributed"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           # CUDA 12 builds
           - build-type: 'cublas'
             cuda-major-version: "12"
@@ -210,6 +249,32 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-fish-speech'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-faster-qwen3-tts'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "faster-qwen3-tts"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "8"
@@ -327,6 +392,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-trl'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "trl"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "8"
@@ -457,6 +535,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-mlx-distributed'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-distributed"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "8"
@@ -483,6 +574,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.golang"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "8"
@@ -575,6 +679,32 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-fish-speech'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-faster-qwen3-tts'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "faster-qwen3-tts"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -666,6 +796,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-trl'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "trl"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'l4t'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -705,6 +848,32 @@ jobs:
             backend: "qwen-tts"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-fish-speech'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-faster-qwen3-tts'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "faster-qwen3-tts"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
           - build-type: 'l4t'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -718,6 +887,19 @@ jobs:
             backend: "pocket-tts"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-chatterbox'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "chatterbox"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
           - build-type: 'l4t'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -770,6 +952,19 @@ jobs:
             backend: "mlx-audio"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-mlx-distributed'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "mlx-distributed"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -874,6 +1069,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-mlx-distributed'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-distributed"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -926,6 +1134,32 @@ jobs:
             backend: "whisper"
             dockerfile: "./backend/Dockerfile.golang"
             context: "./"
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            skip-drivers: 'false'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-acestep-cpp'
+            base-image: "ubuntu:24.04"
+            ubuntu-version: '2404'
+            runs-on: 'ubuntu-24.04-arm'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -1097,6 +1331,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-fish-speech'
+            runs-on: 'arc-runner-set'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            skip-drivers: 'false'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'hipblas'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1293,6 +1540,32 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-fish-speech'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-faster-qwen3-tts'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "faster-qwen3-tts"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
           - build-type: 'l4t'
             cuda-major-version: "12"
             cuda-minor-version: "0"
@@ -1358,6 +1631,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-mlx-distributed'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "mlx-distributed"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
           # SYCL additional backends
           - build-type: 'intel'
             cuda-major-version: ""
@@ -1437,6 +1723,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'intel'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-intel-fish-speech'
+            runs-on: 'arc-runner-set'
+            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
+            skip-drivers: 'false'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'intel'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1529,6 +1828,98 @@ jobs:
             dockerfile: "./backend/Dockerfile.llama-cpp"
             context: "./"
             ubuntu-version: '2404'
+          # llama-cpp-tq (TurboQuant fork)
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-llama-cpp-tq'
+            runs-on: 'bigger-runner'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-llama-cpp-tq'
+            runs-on: 'bigger-runner'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-llama-cpp-tq'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            skip-drivers: 'false'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-llama-cpp-tq'
+            base-image: "ubuntu:24.04"
+            runs-on: 'ubuntu-24.04-arm'
+            ubuntu-version: '2404'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-llama-cpp-tq'
+            runs-on: 'ubuntu-latest'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            skip-drivers: 'false'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            skip-drivers: 'false'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-arm64-llama-cpp-tq'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            runs-on: 'ubuntu-24.04-arm'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2204'
+          - build-type: 'vulkan'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-vulkan-llama-cpp-tq'
+            runs-on: 'bigger-runner'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "llama-cpp-tq"
+            dockerfile: "./backend/Dockerfile.llama-cpp"
+            context: "./"
+            ubuntu-version: '2404'
           # Stablediffusion-ggml
           - build-type: ''
             cuda-major-version: ""
@@ -1674,6 +2065,113 @@ jobs:
             dockerfile: "./backend/Dockerfile.golang"
             context: "./"
             ubuntu-version: '2404'
+          # acestep-cpp
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'sycl_f32'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-intel-sycl-f32-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'sycl_f16'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-intel-sycl-f16-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'vulkan'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-vulkan-acestep-cpp'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            skip-drivers: 'false'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-arm64-acestep-cpp'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            runs-on: 'ubuntu-24.04-arm'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2204'
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-acestep-cpp'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            runs-on: 'ubuntu-latest'
+            skip-drivers: 'false'
+            backend: "acestep-cpp"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          # voxtral
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-voxtral'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "voxtral"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
+          #opus
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-opus'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "opus"
+            dockerfile: "./backend/Dockerfile.golang"
+            context: "./"
+            ubuntu-version: '2404'
           #silero-vad
           - build-type: ''
             cuda-major-version: ""
@@ -1702,20 +2200,6 @@ jobs:
             dockerfile: "./backend/Dockerfile.golang"
             context: "./"
             ubuntu-version: '2404'
-          # huggingface
-          - build-type: ''
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64,linux/arm64'
-            tag-latest: 'auto'
-            tag-suffix: '-huggingface'
-            runs-on: 'ubuntu-latest'
-            base-image: "ubuntu:24.04"
-            skip-drivers: 'false'
-            backend: "huggingface"
-            dockerfile: "./backend/Dockerfile.golang"
-            context: "./"
-            ubuntu-version: '2404'
           # rfdetr
           - build-type: ''
             cuda-major-version: ""
@@ -1875,6 +2359,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-fish-speech'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "fish-speech"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1937,6 +2434,9 @@ jobs:
           - backend: "mlx-audio"
             tag-suffix: "-metal-darwin-arm64-mlx-audio"
             build-type: "mps"
+          - backend: "mlx-distributed"
+            tag-suffix: "-metal-darwin-arm64-mlx-distributed"
+            build-type: "mps"
           - backend: "stablediffusion-ggml"
             tag-suffix: "-metal-darwin-arm64-stablediffusion-ggml"
             build-type: "metal"
@@ -1945,6 +2445,14 @@ jobs:
             tag-suffix: "-metal-darwin-arm64-whisper"
             build-type: "metal"
             lang: "go"
+          - backend: "acestep-cpp"
+            tag-suffix: "-metal-darwin-arm64-acestep-cpp"
+            build-type: "metal"
+            lang: "go"
+          - backend: "voxtral"
+            tag-suffix: "-metal-darwin-arm64-voxtral"
+            build-type: "metal"
+            lang: "go"
           - backend: "vibevoice"
             tag-suffix: "-metal-darwin-arm64-vibevoice"
             build-type: "mps"
@@ -1957,6 +2465,9 @@ jobs:
           - backend: "qwen-tts"
             tag-suffix: "-metal-darwin-arm64-qwen-tts"
             build-type: "mps"
+          - backend: "fish-speech"
+            tag-suffix: "-metal-darwin-arm64-fish-speech"
+            build-type: "mps"
           - backend: "voxcpm"
             tag-suffix: "-metal-darwin-arm64-voxcpm"
             build-type: "mps"
@@ -1994,6 +2505,10 @@ jobs:
             tag-suffix: "-metal-darwin-arm64-piper"
             build-type: "metal"
             lang: "go"
+          - backend: "opus"
+            tag-suffix: "-metal-darwin-arm64-opus"
+            build-type: "metal"
+            lang: "go"
           - backend: "silero-vad"
             tag-suffix: "-metal-darwin-arm64-silero-vad"
             build-type: "metal"
@@ -2002,10 +2517,9 @@ jobs:
             tag-suffix: "-metal-darwin-arm64-local-store"
             build-type: "metal"
             lang: "go"
-          - backend: "huggingface"
-            tag-suffix: "-metal-darwin-arm64-huggingface"
-            build-type: "metal"
-            lang: "go"
+          - backend: "llama-cpp-quantization"
+            tag-suffix: "-metal-darwin-arm64-llama-cpp-quantization"
+            build-type: "mps"
     with:
       backend: ${{ matrix.backend }}
       build-type: ${{ matrix.build-type }}
@@ -2045,7 +2559,7 @@ jobs:
           make protogen-go
           make backends/llama-cpp-darwin
       - name: Upload llama-cpp.tar
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: llama-cpp-tar
           path: backend-images/llama-cpp.tar
@@ -2055,7 +2569,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Download llama-cpp.tar
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: llama-cpp-tar
           path: .
@@ -2071,7 +2585,7 @@ jobs:
           echo "${{ secrets.LOCALAI_REGISTRY_PASSWORD }}" | crane auth login quay.io -u "${{ secrets.LOCALAI_REGISTRY_USERNAME }}" --password-stdin
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             localai/localai-backends
@@ -2084,7 +2598,7 @@ jobs:
             suffix=-metal-darwin-arm64-llama-cpp,onlatest=true
       - name: Docker meta
         id: quaymeta
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/local-ai-backends

.github/workflows/backend_build.yml

@@ -149,7 +149,7 @@ jobs:
       - name: Docker meta
         id: meta
         if: github.event_name != 'pull_request'
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/local-ai-backends
@@ -165,7 +165,7 @@ jobs:
       - name: Docker meta for PR
         id: meta_pull_request
         if: github.event_name == 'pull_request'
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/ci-tests
@@ -188,21 +188,21 @@ jobs:
 
       - name: Login to DockerHub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           username: ${{ secrets.dockerUsername }}
           password: ${{ secrets.dockerPassword }}
 
       - name: Login to Quay.io
         if: ${{ env.quay_username != '' }}
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           registry: quay.io
           username: ${{ secrets.quayUsername }}
           password: ${{ secrets.quayPassword }}
 
       - name: Build and push
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         if: github.event_name != 'pull_request'
         with:
           builder: ${{ steps.buildx.outputs.name }}
@@ -223,7 +223,7 @@ jobs:
           labels: ${{ steps.meta.outputs.labels }}
 
       - name: Build and push (PR)
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         if: github.event_name == 'pull_request'
         with:
           builder: ${{ steps.buildx.outputs.name }}

.github/workflows/backend_build_darwin.yml

@@ -74,7 +74,7 @@ jobs:
           BACKEND=${{ inputs.backend }} BUILD_TYPE=${{ inputs.build-type }} USE_PIP=${{ inputs.use-pip }} make build-darwin-${{ inputs.lang }}-backend
 
       - name: Upload ${{ inputs.backend }}.tar
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: ${{ inputs.backend }}-tar
           path: backend-images/${{ inputs.backend }}.tar
@@ -85,7 +85,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Download ${{ inputs.backend }}.tar
-        uses: actions/download-artifact@v7
+        uses: actions/download-artifact@v8
         with:
           name: ${{ inputs.backend }}-tar
           path: .
@@ -105,7 +105,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             localai/localai-backends
@@ -119,7 +119,7 @@ jobs:
 
       - name: Docker meta
         id: quaymeta
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/local-ai-backends

.github/workflows/build-test.yaml

@@ -37,7 +37,7 @@ jobs:
           make build-launcher-darwin
           ls -liah dist
       - name: Upload macOS launcher artifacts
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: launcher-macos
           path: dist/
@@ -60,7 +60,7 @@ jobs:
           sudo apt-get install golang gcc libgl1-mesa-dev xorg-dev libxkbcommon-dev
           make build-launcher-linux
       - name: Upload Linux launcher artifacts
-        uses: actions/upload-artifact@v6
+        uses: actions/upload-artifact@v7
         with:
           name: launcher-linux
           path: local-ai-launcher-linux.tar.xz

.github/workflows/bump-inference-defaults.yml

@@ -0,0 +1,48 @@
+name: Bump inference defaults
+
+on:
+  schedule:
+    # Run daily at 06:00 UTC
+    - cron: '0 6 * * *'
+  workflow_dispatch: # Allow manual trigger
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  bump:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v6
+
+      - uses: actions/setup-go@v5
+        with:
+          go-version-file: go.mod
+
+      - name: Re-fetch inference defaults
+        run: make generate-force
+
+      - name: Check for changes
+        id: diff
+        run: |
+          if git diff --quiet core/config/inference_defaults.json; then
+            echo "changed=false" >> "$GITHUB_OUTPUT"
+          else
+            echo "changed=true" >> "$GITHUB_OUTPUT"
+          fi
+
+      - name: Create Pull Request
+        if: steps.diff.outputs.changed == 'true'
+        uses: peter-evans/create-pull-request@v8
+        with:
+          commit-message: "chore: bump inference defaults from unsloth"
+          title: "chore: bump inference defaults from unsloth"
+          body: |
+            Auto-generated update of `core/config/inference_defaults.json` from
+            [unsloth's inference_defaults.json](https://github.com/unslothai/unsloth/blob/main/studio/backend/assets/configs/inference_defaults.json).
+
+            This PR was created automatically by the `bump-inference-defaults` workflow.
+          branch: chore/bump-inference-defaults
+          delete-branch: true
+          labels: automated

.github/workflows/bump_deps.yaml

@@ -14,14 +14,15 @@ jobs:
             variable: "LLAMA_VERSION"
             branch: "master"
             file: "backend/cpp/llama-cpp/Makefile"
+          - repository: "TheTom/llama-cpp-turboquant"
+            variable: "LLAMA_VERSION"
+            branch: "master"
+            file: "backend/cpp/llama-cpp-tq/Makefile"
+            branch_suffix: "-tq"
           - repository: "ggml-org/whisper.cpp"
             variable: "WHISPER_CPP_VERSION"
             branch: "master"
             file: "backend/go/whisper/Makefile"
-          - repository: "PABannier/bark.cpp"
-            variable: "BARKCPP_VERSION"
-            branch: "main"
-            file: "Makefile"
           - repository: "leejet/stable-diffusion.cpp"
             variable: "STABLEDIFFUSION_GGML_VERSION"
             branch: "master"
@@ -30,6 +31,14 @@ jobs:
             variable: "PIPER_VERSION"
             branch: "master"
             file: "backend/go/piper/Makefile"
+          - repository: "antirez/voxtral.c"
+            variable: "VOXTRAL_VERSION"
+            branch: "main"
+            file: "backend/go/voxtral/Makefile"
+          - repository: "ace-step/acestep.cpp"
+            variable: "ACESTEP_CPP_VERSION"
+            branch: "master"
+            file: "backend/go/acestep-cpp/Makefile"
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v6
@@ -56,7 +65,7 @@ jobs:
           push-to-fork: ci-forks/LocalAI
           commit-message: ':arrow_up: Update ${{ matrix.repository }}'
           title: 'chore: :arrow_up: Update ${{ matrix.repository }} to `${{ steps.bump.outputs.commit }}`'
-          branch: "update/${{ matrix.variable }}"
+          branch: "update/${{ matrix.variable }}${{ matrix.branch_suffix }}"
           body: ${{ steps.bump.outputs.message }}
           signoff: true
 

.github/workflows/disabled/localaibot_automerge.yml

@@ -10,7 +10,7 @@ permissions:
   actions: write # to dispatch publish workflow
 jobs:
   dependabot:
-    if: github.repository == 'mudler/LocalAI' && github.actor == 'localai-bot' && !contains(github.event.pull_request.title, 'chore(model gallery):')
+    if: github.repository == 'mudler/LocalAI' && github.actor == 'localai-bot' && contains(github.event.pull_request.title, 'chore:')
     runs-on: ubuntu-latest
     steps:
       - name: Checkout repository

.github/workflows/gallery-agent.yaml

@@ -50,12 +50,12 @@ jobs:
           PATH="$PATH:$HOME/go/bin" make protogen-go
       - uses: mudler/localai-github-action@v1.1
         with:
-          model: 'https://huggingface.co/bartowski/Qwen_Qwen3-1.7B-GGUF'
+          model: 'https://huggingface.co/unsloth/Qwen3.5-2B-GGUF'
 
       - name: Run gallery agent
         env:
           #OPENAI_MODEL: ${{ secrets.OPENAI_MODEL }}
-          OPENAI_MODE: Qwen_Qwen3-1.7B-GGUF
+          OPENAI_MODEL: Qwen3.5-2B-GGUF
           OPENAI_BASE_URL: "http://localhost:8080"
           OPENAI_KEY: ${{ secrets.OPENAI_KEY }}
           #OPENAI_BASE_URL: ${{ secrets.OPENAI_BASE_URL }}

.github/workflows/generate_grpc_cache.yaml

@@ -77,7 +77,7 @@ jobs:
         uses: actions/checkout@v6
 
       - name: Cache GRPC
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         with:
           builder: ${{ steps.buildx.outputs.name }}
           # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.

.github/workflows/generate_intel_image.yaml

@@ -27,14 +27,14 @@ jobs:
           platforms: all
       - name: Login to DockerHub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_PASSWORD }}
 
       - name: Login to quay
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           registry: quay.io
           username: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
@@ -47,7 +47,7 @@ jobs:
         uses: actions/checkout@v6
 
       - name: Cache Intel images
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         with:
           builder: ${{ steps.buildx.outputs.name }}
           build-args: |

.github/workflows/gh-pages.yml

@@ -0,0 +1,75 @@
+name: Deploy docs to GitHub Pages
+
+on:
+  push:
+    branches:
+      - master
+    paths:
+      - 'docs/**'
+      - 'gallery/**'
+      - 'images/**'
+      - '.github/ci/modelslist.go'
+      - '.github/workflows/gh-pages.yml'
+  workflow_dispatch:
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+concurrency:
+  group: pages
+  cancel-in-progress: false
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    env:
+      HUGO_VERSION: "0.146.3"
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v6
+        with:
+          fetch-depth: 0  # needed for enableGitInfo
+          submodules: true
+
+      - name: Setup Go
+        uses: actions/setup-go@v5
+        with:
+          go-version: '1.22'
+          cache: false
+
+      - name: Setup Hugo
+        uses: peaceiris/actions-hugo@v3
+        with:
+          hugo-version: ${{ env.HUGO_VERSION }}
+          extended: true
+
+      - name: Setup Pages
+        id: pages
+        uses: actions/configure-pages@v6
+
+      - name: Generate gallery
+        run: go run ./.github/ci/modelslist.go ./gallery/index.yaml > docs/static/gallery.html
+
+      - name: Build site
+        working-directory: docs
+        run: |
+          mkdir -p layouts/_default
+          hugo --minify --baseURL "${{ steps.pages.outputs.base_url }}/"
+
+      - name: Upload artifact
+        uses: actions/upload-pages-artifact@v4
+        with:
+          path: docs/public
+
+  deploy:
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
+    runs-on: ubuntu-latest
+    needs: build
+    steps:
+      - name: Deploy to GitHub Pages
+        id: deployment
+        uses: actions/deploy-pages@v5

.github/workflows/image.yml

@@ -26,7 +26,6 @@
         runs-on: ${{ matrix.runs-on }}
         base-image: ${{ matrix.base-image }}
         grpc-base-image: ${{ matrix.grpc-base-image }}
-        aio: ${{ matrix.aio }}
         makeflags: ${{ matrix.makeflags }}
         ubuntu-version: ${{ matrix.ubuntu-version }}
         ubuntu-codename: ${{ matrix.ubuntu-codename }}
@@ -46,7 +45,6 @@
               grpc-base-image: "ubuntu:24.04"
               runs-on: 'ubuntu-latest'
               makeflags: "--jobs=3 --output-sync=target"
-              aio: "-aio-gpu-hipblas"
               ubuntu-version: '2404'
               ubuntu-codename: 'noble'
   
@@ -61,7 +59,6 @@
         cuda-minor-version: ${{ matrix.cuda-minor-version }}
         platforms: ${{ matrix.platforms }}
         runs-on: ${{ matrix.runs-on }}
-        aio: ${{ matrix.aio }}
         base-image: ${{ matrix.base-image }}
         grpc-base-image: ${{ matrix.grpc-base-image }}
         makeflags: ${{ matrix.makeflags }}
@@ -83,7 +80,6 @@
               tag-suffix: ''
               base-image: "ubuntu:24.04"
               runs-on: 'ubuntu-latest'
-              aio: "-aio-cpu"
               makeflags: "--jobs=4 --output-sync=target"
               skip-drivers: 'false'
               ubuntu-version: '2404'
@@ -98,7 +94,6 @@
               base-image: "ubuntu:24.04"
               skip-drivers: 'false'
               makeflags: "--jobs=4 --output-sync=target"
-              aio: "-aio-gpu-nvidia-cuda-12"
               ubuntu-version: '2404'
               ubuntu-codename: 'noble'
             - build-type: 'cublas'
@@ -111,7 +106,6 @@
               base-image: "ubuntu:22.04"
               skip-drivers: 'false'
               makeflags: "--jobs=4 --output-sync=target"
-              aio: "-aio-gpu-nvidia-cuda-13"
               ubuntu-version: '2404'
               ubuntu-codename: 'noble'
             - build-type: 'vulkan'
@@ -122,7 +116,6 @@
               base-image: "ubuntu:24.04"
               skip-drivers: 'false'
               makeflags: "--jobs=4 --output-sync=target"
-              aio: "-aio-gpu-vulkan"
               ubuntu-version: '2404'
               ubuntu-codename: 'noble'
             - build-type: 'intel'
@@ -133,7 +126,6 @@
               tag-suffix: '-gpu-intel'
               runs-on: 'ubuntu-latest'
               makeflags: "--jobs=3 --output-sync=target"
-              aio: "-aio-gpu-intel"
               ubuntu-version: '2404'
               ubuntu-codename: 'noble'
   
@@ -148,7 +140,6 @@
         cuda-minor-version: ${{ matrix.cuda-minor-version }}
         platforms: ${{ matrix.platforms }}
         runs-on: ${{ matrix.runs-on }}
-        aio: ${{ matrix.aio }}
         base-image: ${{ matrix.base-image }}
         grpc-base-image: ${{ matrix.grpc-base-image }}
         makeflags: ${{ matrix.makeflags }}

.github/workflows/image_build.yml

@@ -51,11 +51,6 @@ on:
         required: false
         default: '--jobs=4 --output-sync=target'
         type: string
-      aio:
-        description: 'AIO Image Name'
-        required: false
-        default: ''
-        type: string
       ubuntu-version:
         description: 'Ubuntu version'
         required: false
@@ -151,7 +146,7 @@ jobs:
       - name: Docker meta
         id: meta
         if: github.event_name != 'pull_request'
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/local-ai
@@ -166,7 +161,7 @@ jobs:
       - name: Docker meta for PR
         id: meta_pull_request
         if: github.event_name == 'pull_request'
-        uses: docker/metadata-action@v5
+        uses: docker/metadata-action@v6
         with:
           images: |
             quay.io/go-skynet/ci-tests
@@ -177,34 +172,6 @@ jobs:
           flavor: |
             latest=${{ inputs.tag-latest }}
             suffix=${{ inputs.tag-suffix }}
-      - name: Docker meta AIO (quay.io)
-        if: inputs.aio != ''
-        id: meta_aio
-        uses: docker/metadata-action@v5
-        with:
-          images: |
-            quay.io/go-skynet/local-ai
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-          flavor: |
-            latest=${{ inputs.tag-latest }}
-            suffix=${{ inputs.aio }},onlatest=true
-
-      - name: Docker meta AIO (dockerhub)
-        if: inputs.aio != ''
-        id: meta_aio_dockerhub
-        uses: docker/metadata-action@v5
-        with:
-          images: |
-            localai/localai
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-          flavor: |
-            latest=${{ inputs.tag-latest }}
-            suffix=${{ inputs.aio }},onlatest=true
-
       - name: Set up QEMU
         uses: docker/setup-qemu-action@master
         with:
@@ -216,21 +183,21 @@ jobs:
 
       - name: Login to DockerHub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           username: ${{ secrets.dockerUsername }}
           password: ${{ secrets.dockerPassword }}
 
       - name: Login to DockerHub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
+        uses: docker/login-action@v4
         with:
           registry: quay.io
           username: ${{ secrets.quayUsername }}
           password: ${{ secrets.quayPassword }}
 
       - name: Build and push
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         if: github.event_name != 'pull_request'
         with:
           builder: ${{ steps.buildx.outputs.name }}
@@ -259,7 +226,7 @@ jobs:
           labels: ${{ steps.meta.outputs.labels }}
 ### Start testing image
       - name: Build and push
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         if: github.event_name == 'pull_request'
         with:
           builder: ${{ steps.buildx.outputs.name }}
@@ -287,41 +254,6 @@ jobs:
           tags: ${{ steps.meta_pull_request.outputs.tags }}
           labels: ${{ steps.meta_pull_request.outputs.labels }}
 ## End testing image
-      - name: Build and push AIO image
-        if: inputs.aio != ''
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BASE_IMAGE=quay.io/go-skynet/local-ai:${{ steps.meta.outputs.version }}
-            MAKEFLAGS=${{ inputs.makeflags }}
-          context: .
-          file: ./Dockerfile.aio
-          platforms: ${{ inputs.platforms }}
-          push: ${{ github.event_name != 'pull_request' }}
-          tags: ${{ steps.meta_aio.outputs.tags }}
-          labels: ${{ steps.meta_aio.outputs.labels }}
-
-      - name: Build and push AIO image (dockerhub)
-        if: inputs.aio != ''
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BASE_IMAGE=localai/localai:${{ steps.meta.outputs.version }}
-            MAKEFLAGS=${{ inputs.makeflags }}
-          context: .
-          file: ./Dockerfile.aio
-          platforms: ${{ inputs.platforms }}
-          push: ${{ github.event_name != 'pull_request' }}
-          tags: ${{ steps.meta_aio_dockerhub.outputs.tags }}
-          labels: ${{ steps.meta_aio_dockerhub.outputs.labels }}
-
       - name: job summary
         run: |
           echo "Built image: ${{ steps.meta.outputs.labels }}" >> $GITHUB_STEP_SUMMARY
-
-      - name: job summary(AIO)
-        if: inputs.aio != ''
-        run: |
-          echo "Built image: ${{ steps.meta_aio.outputs.labels }}" >> $GITHUB_STEP_SUMMARY

.github/workflows/release.yaml

@@ -18,7 +18,7 @@ jobs:
         with:
           go-version: 1.23
       - name: Run GoReleaser
-        uses: goreleaser/goreleaser-action@v6
+        uses: goreleaser/goreleaser-action@v7
         with:
           version: v2.11.0
           args: release --clean

.github/workflows/stalebot.yml

@@ -11,7 +11,7 @@ jobs:
     if: github.repository == 'mudler/LocalAI'
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/stale@997185467fa4f803885201cee163a9f38240193d # v9
+      - uses: actions/stale@b5d41d4e1d5dceea10e7104786b73624c18a190f # v9
         with:
           stale-issue-message: 'This issue is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 5 days.'
           stale-pr-message: 'This PR is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 10 days.'

.github/workflows/test-extra.yml

@@ -14,6 +14,37 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
+  detect-changes:
+    runs-on: ubuntu-latest
+    outputs:
+      run-all: ${{ steps.detect.outputs.run-all }}
+      transformers: ${{ steps.detect.outputs.transformers }}
+      rerankers: ${{ steps.detect.outputs.rerankers }}
+      diffusers: ${{ steps.detect.outputs.diffusers }}
+      coqui: ${{ steps.detect.outputs.coqui }}
+      moonshine: ${{ steps.detect.outputs.moonshine }}
+      pocket-tts: ${{ steps.detect.outputs.pocket-tts }}
+      qwen-tts: ${{ steps.detect.outputs.qwen-tts }}
+      qwen-asr: ${{ steps.detect.outputs.qwen-asr }}
+      nemo: ${{ steps.detect.outputs.nemo }}
+      voxcpm: ${{ steps.detect.outputs.voxcpm }}
+      llama-cpp-quantization: ${{ steps.detect.outputs.llama-cpp-quantization }}
+      acestep-cpp: ${{ steps.detect.outputs.acestep-cpp }}
+      voxtral: ${{ steps.detect.outputs.voxtral }}
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v6
+      - name: Setup Bun
+        uses: oven-sh/setup-bun@v2
+      - name: Install dependencies
+        run: bun add js-yaml @octokit/core
+      - name: Detect changed backends
+        id: detect
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          GITHUB_EVENT_PATH: ${{ github.event_path }}
+        run: bun run scripts/changed-backends.js
+
   # Requires CUDA
   # tests-chatterbox-tts:
   #   runs-on: ubuntu-latest
@@ -37,6 +68,8 @@ jobs:
   #          make --jobs=5 --output-sync=target -C backend/python/chatterbox
   #          make --jobs=5 --output-sync=target -C backend/python/chatterbox test
   tests-transformers:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.transformers == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -58,6 +91,8 @@ jobs:
            make --jobs=5 --output-sync=target -C backend/python/transformers
            make --jobs=5 --output-sync=target -C backend/python/transformers test
   tests-rerankers:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.rerankers == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -80,6 +115,8 @@ jobs:
            make --jobs=5 --output-sync=target -C backend/python/rerankers test
 
   tests-diffusers:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.diffusers == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -229,6 +266,8 @@ jobs:
   #          make --jobs=5 --output-sync=target -C backend/python/vllm test
 
   tests-coqui:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.coqui == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -248,6 +287,8 @@ jobs:
           make --jobs=5 --output-sync=target -C backend/python/coqui
           make --jobs=5 --output-sync=target -C backend/python/coqui test
   tests-moonshine:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.moonshine == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -267,6 +308,8 @@ jobs:
           make --jobs=5 --output-sync=target -C backend/python/moonshine
           make --jobs=5 --output-sync=target -C backend/python/moonshine test
   tests-pocket-tts:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.pocket-tts == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -286,6 +329,8 @@ jobs:
           make --jobs=5 --output-sync=target -C backend/python/pocket-tts
           make --jobs=5 --output-sync=target -C backend/python/pocket-tts test
   tests-qwen-tts:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.qwen-tts == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -304,7 +349,31 @@ jobs:
         run: |
           make --jobs=5 --output-sync=target -C backend/python/qwen-tts
           make --jobs=5 --output-sync=target -C backend/python/qwen-tts test
+  # TODO: s2-pro model is too large to load on CPU-only CI runners — re-enable
+  # when we have GPU runners or a smaller test model.
+  # tests-fish-speech:
+  #   runs-on: ubuntu-latest
+  #   timeout-minutes: 45
+  #   steps:
+  #     - name: Clone
+  #       uses: actions/checkout@v6
+  #       with:
+  #         submodules: true
+  #     - name: Dependencies
+  #       run: |
+  #         sudo apt-get update
+  #         sudo apt-get install -y build-essential ffmpeg portaudio19-dev
+  #         sudo apt-get install -y ca-certificates cmake curl patch python3-pip
+  #         # Install UV
+  #         curl -LsSf https://astral.sh/uv/install.sh | sh
+  #         pip install --user --no-cache-dir grpcio-tools==1.64.1
+  #     - name: Test fish-speech
+  #       run: |
+  #         make --jobs=5 --output-sync=target -C backend/python/fish-speech
+  #         make --jobs=5 --output-sync=target -C backend/python/fish-speech test
   tests-qwen-asr:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.qwen-asr == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -324,6 +393,8 @@ jobs:
           make --jobs=5 --output-sync=target -C backend/python/qwen-asr
           make --jobs=5 --output-sync=target -C backend/python/qwen-asr test
   tests-nemo:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.nemo == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -343,6 +414,8 @@ jobs:
           make --jobs=5 --output-sync=target -C backend/python/nemo
           make --jobs=5 --output-sync=target -C backend/python/nemo test
   tests-voxcpm:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.voxcpm == 'true' || needs.detect-changes.outputs.run-all == 'true'
     runs-on: ubuntu-latest
     steps:
       - name: Clone
@@ -361,3 +434,97 @@ jobs:
         run: |
           make --jobs=5 --output-sync=target -C backend/python/voxcpm
           make --jobs=5 --output-sync=target -C backend/python/voxcpm test
+  tests-llama-cpp-quantization:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.llama-cpp-quantization == 'true' || needs.detect-changes.outputs.run-all == 'true'
+    runs-on: ubuntu-latest
+    timeout-minutes: 30
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential cmake curl git python3-pip
+          # Install UV
+          curl -LsSf https://astral.sh/uv/install.sh | sh
+          pip install --user --no-cache-dir grpcio-tools==1.64.1
+      - name: Build llama-quantize from llama.cpp
+        run: |
+          git clone --depth 1 https://github.com/ggml-org/llama.cpp.git /tmp/llama.cpp
+          cmake -B /tmp/llama.cpp/build -S /tmp/llama.cpp -DGGML_NATIVE=OFF
+          cmake --build /tmp/llama.cpp/build --target llama-quantize -j$(nproc)
+          sudo cp /tmp/llama.cpp/build/bin/llama-quantize /usr/local/bin/
+      - name: Install backend
+        run: |
+          make --jobs=5 --output-sync=target -C backend/python/llama-cpp-quantization
+      - name: Test llama-cpp-quantization
+        run: |
+          make --jobs=5 --output-sync=target -C backend/python/llama-cpp-quantization test
+  tests-acestep-cpp:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.acestep-cpp == 'true' || needs.detect-changes.outputs.run-all == 'true'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential cmake curl libopenblas-dev ffmpeg
+      - name: Setup Go
+        uses: actions/setup-go@v5
+      - name: Display Go version
+        run: go version
+      - name: Proto Dependencies
+        run: |
+          # Install protoc
+          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
+          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
+          rm protoc.zip
+          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
+          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
+          PATH="$PATH:$HOME/go/bin" make protogen-go
+      - name: Build acestep-cpp
+        run: |
+          make --jobs=5 --output-sync=target -C backend/go/acestep-cpp
+      - name: Test acestep-cpp
+        run: |
+          make --jobs=5 --output-sync=target -C backend/go/acestep-cpp test
+  tests-voxtral:
+    needs: detect-changes
+    if: needs.detect-changes.outputs.voxtral == 'true' || needs.detect-changes.outputs.run-all == 'true'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential cmake curl libopenblas-dev ffmpeg
+      - name: Setup Go
+        uses: actions/setup-go@v5
+      # You can test your matrix by printing the current Go version
+      - name: Display Go version
+        run: go version
+      - name: Proto Dependencies
+        run: |
+          # Install protoc
+          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
+          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
+          rm protoc.zip
+          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
+          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
+          PATH="$PATH:$HOME/go/bin" make protogen-go
+      - name: Build voxtral
+        run: |
+          make --jobs=5 --output-sync=target -C backend/go/voxtral
+      - name: Test voxtral
+        run: |
+          make --jobs=5 --output-sync=target -C backend/go/voxtral test

.github/workflows/test.yml

@@ -21,7 +21,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        go-version: ['1.25.x']
+        go-version: ['1.26.x']
     steps:
       - name: Free Disk Space (Ubuntu)
         uses: jlumbroso/free-disk-space@main
@@ -93,30 +93,21 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ccache upx-ucl curl ffmpeg
-          sudo apt-get install -y libgmock-dev clang
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake patch python3-pip unzip
-          sudo apt-get install -y libopencv-dev
-
-          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
-          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
-          rm protoc.zip
-
-          curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
-          sudo dpkg -i cuda-keyring_1.1-1_all.deb
-          sudo apt-get update
-          sudo apt-get install -y cuda-nvcc-${CUDA_VERSION} libcublas-dev-${CUDA_VERSION}
-          export CUDACXX=/usr/local/cuda/bin/nvcc
-          make -C backend/python/transformers
-
-          make backends/huggingface backends/llama-cpp backends/local-store backends/silero-vad backends/piper backends/whisper backends/stablediffusion-ggml
-        env:
-          CUDA_VERSION: 12-4
+          sudo apt-get install curl ffmpeg libopus-dev
+      - name: Setup Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: '22'
+      - name: Build React UI
+        run: make react-ui
+      - name: Build backends
+        run: |
+          make backends/transformers
+          mkdir external && mv backends/transformers external/transformers
+          make backends/llama-cpp backends/local-store backends/silero-vad backends/piper backends/whisper backends/stablediffusion-ggml
       - name: Test
         run: |
-          PATH="$PATH:/root/go/bin" GO_TAGS="tts" make --jobs 5 --output-sync=target test
+          TRANSFORMER_BACKEND=$PWD/external/transformers/run.sh PATH="$PATH:/root/go/bin" GO_TAGS="tts" make --jobs 5 --output-sync=target test
       - name: Setup tmate session if tests fail
         if: ${{ failure() }}
         uses: mxschmitt/action-tmate@v3.23
@@ -125,7 +116,7 @@ jobs:
           connect-timeout-seconds: 180
           limit-access-to-actor: true
 
-  tests-aio-container:
+  tests-e2e-container:
     runs-on: ubuntu-latest
     steps:
       - name: Release space from worker
@@ -175,7 +166,7 @@ jobs:
           PATH="$PATH:$HOME/go/bin" make protogen-go
       - name: Test
         run: |
-            PATH="$PATH:$HOME/go/bin" make backends/local-store backends/silero-vad backends/llama-cpp backends/whisper backends/piper backends/stablediffusion-ggml docker-build-aio e2e-aio
+            PATH="$PATH:$HOME/go/bin" make backends/local-store backends/silero-vad backends/llama-cpp backends/whisper backends/piper backends/stablediffusion-ggml docker-build-e2e e2e-aio
       - name: Setup tmate session if tests fail
         if: ${{ failure() }}
         uses: mxschmitt/action-tmate@v3.23
@@ -188,7 +179,7 @@ jobs:
     runs-on: macos-latest
     strategy:
       matrix:
-        go-version: ['1.25.x']
+        go-version: ['1.26.x']
     steps:
       - name: Clone
         uses: actions/checkout@v6
@@ -204,8 +195,14 @@ jobs:
         run: go version
       - name: Dependencies
         run: |
-          brew install protobuf grpc make protoc-gen-go protoc-gen-go-grpc libomp llvm
+          brew install protobuf grpc make protoc-gen-go protoc-gen-go-grpc libomp llvm opus
           pip install --user --no-cache-dir grpcio-tools grpcio
+      - name: Setup Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: '22'
+      - name: Build React UI
+        run: make react-ui
       - name: Build llama-cpp-darwin
         run: |
           make protogen-go

.github/workflows/tests-e2e.yml

@@ -43,7 +43,13 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install -y build-essential
+          sudo apt-get install -y build-essential libopus-dev
+      - name: Setup Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: '22'
+      - name: Build React UI
+        run: make react-ui
       - name: Test Backend E2E
         run: |
           PATH="$PATH:$HOME/go/bin" make build-mock-backend test-e2e

.github/workflows/tests-ui-e2e.yml

@@ -0,0 +1,72 @@
+---
+name: 'UI E2E Tests'
+
+on:
+  pull_request:
+    paths:
+      - 'core/http/**'
+      - 'tests/e2e-ui/**'
+      - 'tests/e2e/mock-backend/**'
+  push:
+    branches:
+      - master
+
+concurrency:
+  group: ci-tests-ui-e2e-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
+jobs:
+  tests-ui-e2e:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        go-version: ['1.26.x']
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Setup Go ${{ matrix.go-version }}
+        uses: actions/setup-go@v5
+        with:
+          go-version: ${{ matrix.go-version }}
+          cache: false
+      - name: Setup Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: '22'
+      - name: Proto Dependencies
+        run: |
+          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
+          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
+          rm protoc.zip
+          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
+          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
+      - name: System Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential libopus-dev
+      - name: Build UI test server
+        run: PATH="$PATH:$HOME/go/bin" make build-ui-test-server
+      - name: Install Playwright
+        working-directory: core/http/react-ui
+        run: |
+          npm install
+          npx playwright install --with-deps chromium
+      - name: Run Playwright tests
+        working-directory: core/http/react-ui
+        run: npx playwright test
+      - name: Upload Playwright report
+        if: ${{ failure() }}
+        uses: actions/upload-artifact@v7
+        with:
+          name: playwright-report
+          path: core/http/react-ui/playwright-report/
+          retention-days: 7
+      - name: Setup tmate session if tests fail
+        if: ${{ failure() }}
+        uses: mxschmitt/action-tmate@v3.23
+        with:
+          detached: true
+          connect-timeout-seconds: 180
+          limit-access-to-actor: true

.gitignore

@@ -9,6 +9,7 @@ prepare-sources
 /backend/cpp/llama-cpp/llama.cpp
 /backend/cpp/llama-*
 !backend/cpp/llama-cpp
+!backend/cpp/llama-cpp-tq
 /backends
 /backend-images
 /result.yaml
@@ -37,7 +38,7 @@ models/*
 test-models/
 test-dir/
 tests/e2e-aio/backends
-tests/e2e-aio/models
+mock-backend
 
 release/
 
@@ -65,3 +66,15 @@ docs/static/gallery.html
 
 # per-developer customization files for the development container
 .devcontainer/customization/*
+
+# React UI build artifacts (keep placeholder dist/index.html)
+core/http/react-ui/node_modules/
+core/http/react-ui/dist
+
+# Extracted backend binaries for container-based testing
+local-backends/
+
+# UI E2E test artifacts
+tests/e2e-ui/ui-test-server
+core/http/react-ui/playwright-report/
+core/http/react-ui/test-results/

.goreleaser.yaml

@@ -2,6 +2,7 @@ version: 2
 before:
   hooks:
     - make protogen-go
+    - make react-ui
     - go mod tidy
 dist: release
 source:

AGENTS.md

@@ -1,290 +1,24 @@
-# Build and testing
-
-Building and testing the project depends on the components involved and the platform where development is taking place. Due to the amount of context required it's usually best not to try building or testing the project unless the user requests it. If you must build the project then inspect the Makefile in the project root and the Makefiles of any backends that are effected by changes you are making. In addition the workflows in .github/workflows can be used as a reference when it is unclear how to build or test a component. The primary Makefile contains targets for building inside or outside Docker, if the user has not previously specified a preference then ask which they would like to use.
-
-## Building a specified backend
-
-Let's say the user wants to build a particular backend for a given platform. For example let's say they want to build coqui for ROCM/hipblas
-
-- The Makefile has targets like `docker-build-coqui` created with `generate-docker-build-target` at the time of writing. Recently added backends may require a new target.
-- At a minimum we need to set the BUILD_TYPE, BASE_IMAGE build-args
-  - Use .github/workflows/backend.yml as a reference it lists the needed args in the `include` job strategy matrix
-  - l4t and cublas also requires the CUDA major and minor version
-- You can pretty print a command like `DOCKER_MAKEFLAGS=-j$(nproc --ignore=1) BUILD_TYPE=hipblas BASE_IMAGE=rocm/dev-ubuntu-24.04:6.4.4 make docker-build-coqui`
-- Unless the user specifies that they want you to run the command, then just print it because not all agent frontends handle long running jobs well and the output may overflow your context
-- The user may say they want to build AMD or ROCM instead of hipblas, or Intel instead of SYCL or NVIDIA insted of l4t or cublas. Ask for confirmation if there is ambiguity.
-- Sometimes the user may need extra parameters to be added to `docker build` (e.g. `--platform` for cross-platform builds or `--progress` to view the full logs), in which case you can generate the `docker build` command directly.
-
-## Adding a New Backend
-
-When adding a new backend to LocalAI, you need to update several files to ensure the backend is properly built, tested, and registered. Here's a step-by-step guide based on the pattern used for adding backends like `moonshine`:
-
-### 1. Create Backend Directory Structure
-
-Create the backend directory under the appropriate location:
-- **Python backends**: `backend/python/<backend-name>/`
-- **Go backends**: `backend/go/<backend-name>/`
-- **C++ backends**: `backend/cpp/<backend-name>/`
-
-For Python backends, you'll typically need:
-- `backend.py` - Main gRPC server implementation
-- `Makefile` - Build configuration
-- `install.sh` - Installation script for dependencies
-- `protogen.sh` - Protocol buffer generation script
-- `requirements.txt` - Python dependencies
-- `run.sh` - Runtime script
-- `test.py` / `test.sh` - Test files
-
-### 2. Add Build Configurations to `.github/workflows/backend.yml`
-
-Add build matrix entries for each platform/GPU type you want to support. Look at similar backends (e.g., `chatterbox`, `faster-whisper`) for reference.
-
-**Placement in file:**
-- CPU builds: Add after other CPU builds (e.g., after `cpu-chatterbox`)
-- CUDA 12 builds: Add after other CUDA 12 builds (e.g., after `gpu-nvidia-cuda-12-chatterbox`)
-- CUDA 13 builds: Add after other CUDA 13 builds (e.g., after `gpu-nvidia-cuda-13-chatterbox`)
-
-**Additional build types you may need:**
-- ROCm/HIP: Use `build-type: 'hipblas'` with `base-image: "rocm/dev-ubuntu-24.04:6.4.4"`
-- Intel/SYCL: Use `build-type: 'intel'` or `build-type: 'sycl_f16'`/`sycl_f32` with `base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"`
-- L4T (ARM): Use `build-type: 'l4t'` with `platforms: 'linux/arm64'` and `runs-on: 'ubuntu-24.04-arm'`
-
-### 3. Add Backend Metadata to `backend/index.yaml`
-
-**Step 3a: Add Meta Definition**
-
-Add a YAML anchor definition in the `## metas` section (around line 2-300). Look for similar backends to use as a template such as `diffusers` or `chatterbox`
-
-**Step 3b: Add Image Entries**
-
-Add image entries at the end of the file, following the pattern of similar backends such as `diffusers` or `chatterbox`. Include both `latest` (production) and `master` (development) tags.
-
-### 4. Update the Makefile
-
-The Makefile needs to be updated in several places to support building and testing the new backend:
-
-**Step 4a: Add to `.NOTPARALLEL`**
-
-Add `backends/<backend-name>` to the `.NOTPARALLEL` line (around line 2) to prevent parallel execution conflicts:
-
-```makefile
-.NOTPARALLEL: ... backends/<backend-name>
-```
-
-**Step 4b: Add to `prepare-test-extra`**
-
-Add the backend to the `prepare-test-extra` target (around line 312) to prepare it for testing:
-
-```makefile
-prepare-test-extra: protogen-python
-	...
-	$(MAKE) -C backend/python/<backend-name>
-```
-
-**Step 4c: Add to `test-extra`**
-
-Add the backend to the `test-extra` target (around line 319) to run its tests:
-
-```makefile
-test-extra: prepare-test-extra
-	...
-	$(MAKE) -C backend/python/<backend-name> test
-```
-
-**Step 4d: Add Backend Definition**
-
-Add a backend definition variable in the backend definitions section (around line 428-457). The format depends on the backend type:
-
-**For Python backends with root context** (like `faster-whisper`, `coqui`):
-```makefile
-BACKEND_<BACKEND_NAME> = <backend-name>|python|.|false|true
-```
-
-**For Python backends with `./backend` context** (like `chatterbox`, `moonshine`):
-```makefile
-BACKEND_<BACKEND_NAME> = <backend-name>|python|./backend|false|true
-```
-
-**For Go backends**:
-```makefile
-BACKEND_<BACKEND_NAME> = <backend-name>|golang|.|false|true
-```
-
-**Step 4e: Generate Docker Build Target**
-
-Add an eval call to generate the docker-build target (around line 480-501):
-
-```makefile
-$(eval $(call generate-docker-build-target,$(BACKEND_<BACKEND_NAME>)))
-```
-
-**Step 4f: Add to `docker-build-backends`**
-
-Add `docker-build-<backend-name>` to the `docker-build-backends` target (around line 507):
-
-```makefile
-docker-build-backends: ... docker-build-<backend-name>
-```
-
-**Determining the Context:**
-
-- If the backend is in `backend/python/<backend-name>/` and uses `./backend` as context in the workflow file, use `./backend` context
-- If the backend is in `backend/python/<backend-name>/` but uses `.` as context in the workflow file, use `.` context
-- Check similar backends to determine the correct context
-
-### 5. Verification Checklist
-
-After adding a new backend, verify:
-
-- [ ] Backend directory structure is complete with all necessary files
-- [ ] Build configurations added to `.github/workflows/backend.yml` for all desired platforms
-- [ ] Meta definition added to `backend/index.yaml` in the `## metas` section
-- [ ] Image entries added to `backend/index.yaml` for all build variants (latest + development)
-- [ ] Tag suffixes match between workflow file and index.yaml
-- [ ] Makefile updated with all 6 required changes (`.NOTPARALLEL`, `prepare-test-extra`, `test-extra`, backend definition, docker-build target eval, `docker-build-backends`)
-- [ ] No YAML syntax errors (check with linter)
-- [ ] No Makefile syntax errors (check with linter)
-- [ ] Follows the same pattern as similar backends (e.g., if it's a transcription backend, follow `faster-whisper` pattern)
-
-### 6. Example: Adding a Python Backend
-
-For reference, when `moonshine` was added:
-- **Files created**: `backend/python/moonshine/{backend.py, Makefile, install.sh, protogen.sh, requirements.txt, run.sh, test.py, test.sh}`
-- **Workflow entries**: 3 build configurations (CPU, CUDA 12, CUDA 13)
-- **Index entries**: 1 meta definition + 6 image entries (cpu, cuda12, cuda13 × latest/development)
-- **Makefile updates**: 
-  - Added to `.NOTPARALLEL` line
-  - Added to `prepare-test-extra` and `test-extra` targets
-  - Added `BACKEND_MOONSHINE = moonshine|python|./backend|false|true`
-  - Added eval for docker-build target generation
-  - Added `docker-build-moonshine` to `docker-build-backends`
-
-# Coding style
-
-- The project has the following .editorconfig
-
-```
-root = true
-
-[*]
-indent_style = space
-indent_size = 2
-end_of_line = lf
-charset = utf-8
-trim_trailing_whitespace = true
-insert_final_newline = true
-
-[*.go]
-indent_style = tab
-
-[Makefile]
-indent_style = tab
-
-[*.proto]
-indent_size = 2
-
-[*.py]
-indent_size = 4
-
-[*.js]
-indent_size = 2
-
-[*.yaml]
-indent_size = 2
-
-[*.md]
-trim_trailing_whitespace = false
-```
-
-- Use comments sparingly to explain why code does something, not what it does. Comments are there to add context that would be difficult to deduce from reading the code.
-- Prefer modern Go e.g. use `any` not `interface{}`
-
-# Logging
-
-Use `github.com/mudler/xlog` for logging which has the same API as slog.
-
-# llama.cpp Backend
-
-The llama.cpp backend (`backend/cpp/llama-cpp/grpc-server.cpp`) is a gRPC adaptation of the upstream HTTP server (`llama.cpp/tools/server/server.cpp`). It uses the same underlying server infrastructure from `llama.cpp/tools/server/server-context.cpp`.
-
-## Building and Testing
-
-- Test llama.cpp backend compilation: `make backends/llama-cpp`
-- The backend is built as part of the main build process
-- Check `backend/cpp/llama-cpp/Makefile` for build configuration
-
-## Architecture
-
-- **grpc-server.cpp**: gRPC server implementation, adapts HTTP server patterns to gRPC
-- Uses shared server infrastructure: `server-context.cpp`, `server-task.cpp`, `server-queue.cpp`, `server-common.cpp`
-- The gRPC server mirrors the HTTP server's functionality but uses gRPC instead of HTTP
-
-## Common Issues When Updating llama.cpp
-
-When fixing compilation errors after upstream changes:
-1. Check how `server.cpp` (HTTP server) handles the same change
-2. Look for new public APIs or getter methods
-3. Store copies of needed data instead of accessing private members
-4. Update function calls to match new signatures
-5. Test with `make backends/llama-cpp`
-
-## Key Differences from HTTP Server
-
-- gRPC uses `BackendServiceImpl` class with gRPC service methods
-- HTTP server uses `server_routes` with HTTP handlers
-- Both use the same `server_context` and task queue infrastructure
-- gRPC methods: `LoadModel`, `Predict`, `PredictStream`, `Embedding`, `Rerank`, `TokenizeString`, `GetMetrics`, `Health`
-
-## Tool Call Parsing Maintenance
-
-When working on JSON/XML tool call parsing functionality, always check llama.cpp for reference implementation and updates:
-
-### Checking for XML Parsing Changes
-
-1. **Review XML Format Definitions**: Check `llama.cpp/common/chat-parser-xml-toolcall.h` for `xml_tool_call_format` struct changes
-2. **Review Parsing Logic**: Check `llama.cpp/common/chat-parser-xml-toolcall.cpp` for parsing algorithm updates
-3. **Review Format Presets**: Check `llama.cpp/common/chat-parser.cpp` for new XML format presets (search for `xml_tool_call_format form`)
-4. **Review Model Lists**: Check `llama.cpp/common/chat.h` for `COMMON_CHAT_FORMAT_*` enum values that use XML parsing:
-   - `COMMON_CHAT_FORMAT_GLM_4_5`
-   - `COMMON_CHAT_FORMAT_MINIMAX_M2`
-   - `COMMON_CHAT_FORMAT_KIMI_K2`
-   - `COMMON_CHAT_FORMAT_QWEN3_CODER_XML`
-   - `COMMON_CHAT_FORMAT_APRIEL_1_5`
-   - `COMMON_CHAT_FORMAT_XIAOMI_MIMO`
-   - Any new formats added
-
-### Model Configuration Options
-
-Always check `llama.cpp` for new model configuration options that should be supported in LocalAI:
-
-1. **Check Server Context**: Review `llama.cpp/tools/server/server-context.cpp` for new parameters
-2. **Check Chat Params**: Review `llama.cpp/common/chat.h` for `common_chat_params` struct changes
-3. **Check Server Options**: Review `llama.cpp/tools/server/server.cpp` for command-line argument changes
-4. **Examples of options to check**:
-   - `ctx_shift` - Context shifting support
-   - `parallel_tool_calls` - Parallel tool calling
-   - `reasoning_format` - Reasoning format options
-   - Any new flags or parameters
-
-### Implementation Guidelines
-
-1. **Feature Parity**: Always aim for feature parity with llama.cpp's implementation
-2. **Test Coverage**: Add tests for new features matching llama.cpp's behavior
-3. **Documentation**: Update relevant documentation when adding new formats or options
-4. **Backward Compatibility**: Ensure changes don't break existing functionality
-
-### Files to Monitor
-
-- `llama.cpp/common/chat-parser-xml-toolcall.h` - Format definitions
-- `llama.cpp/common/chat-parser-xml-toolcall.cpp` - Parsing logic
-- `llama.cpp/common/chat-parser.cpp` - Format presets and model-specific handlers
-- `llama.cpp/common/chat.h` - Format enums and parameter structures
-- `llama.cpp/tools/server/server-context.cpp` - Server configuration options
-
-# Documentation
-
-The project documentation is located in `docs/content`. When adding new features or changing existing functionality, it is crucial to update the documentation to reflect these changes. This helps users understand how to use the new capabilities and ensures the documentation stays relevant.
-
-- **Feature Documentation**: If you add a new feature (like a new backend or API endpoint), create a new markdown file in `docs/content/features/` explaining what it is, how to configure it, and how to use it.
-- **Configuration**: If you modify configuration options, update the relevant sections in `docs/content/`.
-- **Examples**: providing concrete examples (like YAML configuration blocks) is highly encouraged to help users get started quickly.
+# LocalAI Agent Instructions
+
+This file is an index to detailed topic guides in the `.agents/` directory. Read the relevant file(s) for the task at hand — you don't need to load all of them.
+
+## Topics
+
+| File | When to read |
+|------|-------------|
+| [.agents/building-and-testing.md](.agents/building-and-testing.md) | Building the project, running tests, Docker builds for specific platforms |
+| [.agents/adding-backends.md](.agents/adding-backends.md) | Adding a new backend (Python, Go, or C++) — full step-by-step checklist |
+| [.agents/coding-style.md](.agents/coding-style.md) | Code style, editorconfig, logging, documentation conventions |
+| [.agents/llama-cpp-backend.md](.agents/llama-cpp-backend.md) | Working on the llama.cpp backend — architecture, updating, tool call parsing |
+| [.agents/testing-mcp-apps.md](.agents/testing-mcp-apps.md) | Testing MCP Apps (interactive tool UIs) in the React UI |
+| [.agents/api-endpoints-and-auth.md](.agents/api-endpoints-and-auth.md) | Adding API endpoints, auth middleware, feature permissions, user access control |
+| [.agents/debugging-backends.md](.agents/debugging-backends.md) | Debugging runtime backend failures, dependency conflicts, rebuilding backends |
+
+## Quick Reference
+
+- **Logging**: Use `github.com/mudler/xlog` (same API as slog)
+- **Go style**: Prefer `any` over `interface{}`
+- **Comments**: Explain *why*, not *what*
+- **Docs**: Update `docs/content/` when adding features or changing config
+- **Build**: Inspect `Makefile` and `.github/workflows/` — ask the user before running long builds
+- **UI**: The active UI is the React app in `core/http/react-ui/`. The older Alpine.js/HTML UI in `core/http/static/` is pending deprecation — all new UI work goes in the React UI

CLAUDE.md

@@ -0,0 +1 @@
+AGENTS.md

CONTRIBUTING.md

@@ -7,8 +7,10 @@ Thank you for your interest in contributing to LocalAI! We appreciate your time
 - [Getting Started](#getting-started)
   - [Prerequisites](#prerequisites)
   - [Setting up the Development Environment](#setting-up-the-development-environment)
+  - [Environment Variables](#environment-variables)
 - [Contributing](#contributing)
   - [Submitting an Issue](#submitting-an-issue)
+  - [Development Workflow](#development-workflow)
   - [Creating a Pull Request (PR)](#creating-a-pull-request-pr)
 - [Coding Guidelines](#coding-guidelines)
 - [Testing](#testing)
@@ -19,18 +21,122 @@ Thank you for your interest in contributing to LocalAI! We appreciate your time
 
 ### Prerequisites
 
-- Golang [1.21]
-- Git
-- macOS/Linux
+- **Go 1.21+** (the project currently uses Go 1.26 in `go.mod`, but 1.21 is the minimum supported version)
+  - [Download Go](https://go.dev/dl/) or install via your package manager
+  - macOS: `brew install go`
+  - Ubuntu/Debian: follow the [official instructions](https://go.dev/doc/install) (the `apt` version is often outdated)
+  - Verify: `go version`
+- **Git**
+- **GNU Make**
+- **GCC / C/C++ toolchain** (required for CGo and native backends)
+- **Protocol Buffers compiler** (`protoc`) — needed for gRPC code generation
 
-### Setting up the Development Environment and running localAI in the local environment
+#### System dependencies by platform
 
-1. Clone the repository: `git clone https://github.com/go-skynet/LocalAI.git`
-2. Navigate to the project directory: `cd LocalAI`
-3. Install the required dependencies ( see https://localai.io/basics/build/#build-localai-locally )
-4. Build LocalAI: `make build`
-5. Run LocalAI: `./local-ai`
-6. To Build and live reload: `make build-dev`
+<details>
+<summary><strong>Ubuntu / Debian</strong></summary>
+
+```bash
+sudo apt-get update
+sudo apt-get install -y build-essential gcc g++ cmake git wget \
+  protobuf-compiler libprotobuf-dev pkg-config \
+  libopencv-dev libgrpc-dev
+```
+
+</details>
+
+<details>
+<summary><strong>CentOS / RHEL / Fedora</strong></summary>
+
+```bash
+sudo dnf groupinstall -y "Development Tools"
+sudo dnf install -y cmake git wget protobuf-compiler protobuf-devel \
+  opencv-devel grpc-devel
+```
+
+</details>
+
+<details>
+<summary><strong>macOS</strong></summary>
+
+```bash
+xcode-select --install
+brew install cmake git protobuf grpc opencv wget
+```
+
+</details>
+
+<details>
+<summary><strong>Windows</strong></summary>
+
+Use [WSL 2](https://learn.microsoft.com/en-us/windows/wsl/install) with an Ubuntu distribution, then follow the Ubuntu instructions above.
+
+</details>
+
+### Setting up the Development Environment
+
+1. **Clone the repository:**
+
+   ```bash
+   git clone https://github.com/mudler/LocalAI.git
+   cd LocalAI
+   ```
+
+2. **Build LocalAI:**
+
+   ```bash
+   make build
+   ```
+
+   This runs protobuf generation, installs Go tools, builds the React UI, and compiles the `local-ai` binary. Key build variables you can set:
+
+   | Variable | Description | Example |
+   |---|---|---|
+   | `BUILD_TYPE` | GPU/accelerator type (`cublas`, `hipblas`, `intel`, ``) | `BUILD_TYPE=cublas make build` |
+   | `GO_TAGS` | Additional Go build tags | `GO_TAGS=debug make build` |
+   | `CUDA_MAJOR_VERSION` | CUDA major version (default: `13`) | `CUDA_MAJOR_VERSION=12` |
+
+3. **Run LocalAI:**
+
+   ```bash
+   ./local-ai
+   ```
+
+4. **Development mode with live reload:**
+
+   ```bash
+   make build-dev
+   ```
+
+   This installs [`air`](https://github.com/air-verse/air) automatically and watches for file changes, rebuilding and restarting the server on each save.
+
+5. **Containerized build** (no local toolchain needed):
+
+   ```bash
+   make docker
+   ```
+
+   For GPU-specific Docker builds, see the `docker-build-*` targets in the Makefile and refer to [CLAUDE.md](CLAUDE.md) for detailed backend build instructions.
+
+### Environment Variables
+
+LocalAI is configured primarily through environment variables (or equivalent CLI flags). The most useful ones for development are:
+
+| Variable | Description | Default |
+|---|---|---|
+| `LOCALAI_DEBUG` | Enable debug mode | `false` |
+| `LOCALAI_LOG_LEVEL` | Log verbosity (`error`, `warn`, `info`, `debug`, `trace`) | — |
+| `LOCALAI_LOG_FORMAT` | Log format (`default`, `text`, `json`) | `default` |
+| `LOCALAI_MODELS_PATH` | Path to model files | `./models` |
+| `LOCALAI_BACKENDS_PATH` | Path to backend binaries | `./backends` |
+| `LOCALAI_CONFIG_DIR` | Directory for dynamic config files (API keys, external backends) | `./configuration` |
+| `LOCALAI_THREADS` | Number of threads for inference | — |
+| `LOCALAI_ADDRESS` | Bind address for the API server | `:8080` |
+| `LOCALAI_API_KEY` | API key(s) for authentication | — |
+| `LOCALAI_CORS` | Enable CORS | `false` |
+| `LOCALAI_DISABLE_WEBUI` | Disable the web UI | `false` |
+
+See `core/cli/run.go` for the full list of supported environment variables.
 
 ## Contributing
 
@@ -40,43 +146,128 @@ We welcome contributions from everyone! To get started, follow these steps:
 
 If you find a bug, have a feature request, or encounter any issues, please check the [issue tracker](https://github.com/go-skynet/LocalAI/issues) to see if a similar issue has already been reported. If not, feel free to [create a new issue](https://github.com/go-skynet/LocalAI/issues/new) and provide as much detail as possible.
 
-### Creating a Pull Request (PR)
+### Development Workflow
+
+#### Branch naming conventions
+
+Use a descriptive branch name that indicates the type and scope of the change:
+
+- `feature/<short-description>` — new functionality
+- `fix/<short-description>` — bug fixes
+- `docs/<short-description>` — documentation changes
+- `refactor/<short-description>` — code refactoring
+
+#### Commit messages
+
+- Use a short, imperative subject line (e.g., "feat: add whisper backend support", not "Added whisper backend support")
+- Keep the subject under 72 characters
+- Use the body to explain **why** the change was made when the subject alone is not sufficient
+- Use [conventional commits](https://www.conventionalcommits.org/en/v1.0.0/)
+
+#### Creating a Pull Request (PR)
+
+Before jumping into a PR for a massive feature or big change, it is preferred to discuss it first via an issue.
 
 1. Fork the repository.
-2. Create a new branch with a descriptive name: `git checkout -b [branch name]`
-3. Make your changes and commit them.
-4. Push the changes to your fork: `git push origin [branch name]`
-5. Create a new pull request from your branch to the main project's `main` or `master` branch.
-6. Provide a clear description of your changes in the pull request.
-7. Make any requested changes during the review process.
-8. Once your PR is approved, it will be merged into the main project.
+2. Create a new branch: `git checkout -b feature/my-change`
+3. Make your changes, keeping commits focused and atomic.
+4. Run tests locally before pushing (see [Testing](#testing) below).
+5. Push to your fork: `git push origin feature/my-change`
+6. Open a pull request against the `master` branch.
+7. Fill in the PR description with:
+   - What the change does and why
+   - How it was tested
+   - Any breaking changes or migration steps
+8. Respond to review feedback promptly. Push follow-up commits rather than force-pushing amended commits so reviewers can see incremental changes.
+9. Once approved, a maintainer will merge your PR.
 
 ## Coding Guidelines
 
-- No specific coding guidelines at the moment. Please make sure the code can be tested. The most popular lint tools like [`golangci-lint`](https://golangci-lint.run) can help you here.
+This project uses an [`.editorconfig`](.editorconfig) file to define formatting standards (indentation, line endings, charset, etc.). Please configure your editor to respect it.
+
+For AI-assisted development, see [`CLAUDE.md`](CLAUDE.md) for agent-specific guidelines including build instructions and backend architecture details.
+
+### General Principles
+
+- Write code that can be tested. All new features and bug fixes should include test coverage.
+- Use comments sparingly to explain **why** code does something, not **what** it does. Comments should add context that would be difficult to deduce from reading the code alone.
+- Keep changes focused. Avoid unrelated refactors, formatting changes, or feature additions in the same PR.
+
+### Go Code
+
+- Prefer modern Go idioms — for example, use `any` instead of `interface{}`.
+- Use [`golangci-lint`](https://golangci-lint.run) to catch common issues before submitting a PR.
+- Use [`github.com/mudler/xlog`](https://github.com/mudler/xlog) for logging (same API as `slog`). Do not use `fmt.Println` or the standard `log` package for operational logging.
+- Use tab indentation for Go files (as defined in `.editorconfig`).
+
+### Python Code
+
+- Use 4-space indentation (as defined in `.editorconfig`).
+- Include a `requirements.txt` for any new dependencies.
+
+### Code Review
+
+- All contributions go through code review via pull requests.
+- Reviewers will check for correctness, test coverage, adherence to these guidelines, and clarity of intent.
+- Be responsive to review feedback and keep discussions constructive.
 
 ## Testing
 
-`make test` cannot handle all the model now. Please be sure to add a test case for the new features or the part was changed.
+All new features and bug fixes should include test coverage. The project uses [Ginkgo](https://onsi.github.io/ginkgo/) as its test framework.
 
-### Running AIO tests
+### Running unit tests
 
-All-In-One images has a set of tests that automatically verifies that most of the endpoints works correctly, a flow can be :
+```bash
+make test
+```
+
+This downloads test model fixtures, runs protobuf generation, and executes the full test suite including llama-gguf, TTS, and stable-diffusion tests. Note: some tests require model files to be downloaded, so the first run may take longer.
+
+To run tests for a specific package:
+
+```bash
+go test ./core/config/...
+go test ./pkg/model/...
+```
+
+To run a specific test by name using Ginkgo's `--focus` flag:
+
+```bash
+go run github.com/onsi/ginkgo/v2/ginkgo --focus="should load a model" -v -r ./core/
+```
+
+### Running end-to-end tests
+
+The e2e tests run LocalAI in a Docker container and exercise the API:
+
+```bash
+make test-e2e
+```
+
+### Running E2E container tests
+
+These tests build a standard LocalAI Docker image and run it with pre-configured model configs to verify that most endpoints work correctly:
 
 ```bash
 # Build the LocalAI docker image
-make DOCKER_IMAGE=local-ai docker
+make docker-build-e2e
 
-# Build the corresponding AIO image
-BASE_IMAGE=local-ai DOCKER_AIO_IMAGE=local-ai-aio:test make docker-aio
+# Run the e2e tests (uses model configs from tests/e2e-aio/models/)
+make e2e-aio
+```
 
-# Run the AIO e2e tests
-LOCALAI_IMAGE_TAG=test LOCALAI_IMAGE=local-ai-aio make run-e2e-aio
+### Testing backends
+
+To prepare and test extra (Python) backends:
+
+```bash
+make prepare-test-extra   # build Python backends for testing
+make test-extra           # run backend-specific tests
 ```
 
 ## Documentation
 
-We are welcome the contribution of the documents, please open new PR or create a new issue. The documentation is available under `docs/` https://github.com/mudler/LocalAI/tree/master/docs
+We welcome contributions to the documentation. Please open a new PR or create a new issue. The documentation is available under `docs/` https://github.com/mudler/LocalAI/tree/master/docs
 
 ### Gallery YAML Schema
 

Dockerfile

@@ -10,7 +10,7 @@ ENV DEBIAN_FRONTEND=noninteractive
 RUN apt-get update && \
     apt-get install -y --no-install-recommends \
         ca-certificates curl wget espeak-ng libgomp1 \
-        ffmpeg libopenblas0 libopenblas-dev sox && \
+        ffmpeg libopenblas0 libopenblas-dev libopus0 sox && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
@@ -176,7 +176,7 @@ ENV PATH=/opt/rocm/bin:${PATH}
 # The requirements-core target is common to all images.  It should not be placed in requirements-core unless every single build will use it.
 FROM requirements-drivers AS build-requirements
 
-ARG GO_VERSION=1.25.4
+ARG GO_VERSION=1.26.0
 ARG CMAKE_VERSION=3.31.10
 ARG CMAKE_FROM_SOURCE=false
 ARG TARGETARCH
@@ -190,6 +190,7 @@ RUN apt-get update && \
         curl libssl-dev \
         git \
         git-lfs \
+        libopus-dev pkg-config \
         unzip upx-ucl python3 python-is-python3 && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
@@ -255,7 +256,7 @@ RUN apt-get update && \
 
 FROM build-requirements AS builder-base
 
-ARG GO_TAGS=""
+ARG GO_TAGS="auth"
 ARG GRPC_BACKENDS
 ARG MAKEFLAGS
 ARG LD_FLAGS="-s -w"
@@ -291,6 +292,17 @@ EOT
 ###################################
 ###################################
 
+# Build React UI
+FROM node:25-slim AS react-ui-builder
+WORKDIR /app
+COPY core/http/react-ui/package*.json ./
+RUN npm install
+COPY core/http/react-ui/ ./
+RUN npm run build
+
+###################################
+###################################
+
 # Compile backends first in a separate stage
 FROM builder-base AS builder-backends
 ARG TARGETARCH
@@ -307,7 +319,6 @@ COPY ./.git ./.git
 # Some of the Go backends use libs from the main src, we could further optimize the caching by building the CPP backends before here
 COPY ./pkg/grpc ./pkg/grpc
 COPY ./pkg/utils ./pkg/utils
-COPY ./pkg/langchain ./pkg/langchain
 
 RUN ls -l ./
 RUN make protogen-go
@@ -320,6 +331,9 @@ WORKDIR /build
 
 COPY . .
 
+# Copy pre-built React UI
+COPY --from=react-ui-builder /app/dist ./core/http/react-ui/dist
+
 ## Build the binary
 ## If we're on arm64 AND using cublas/hipblas, skip some of the llama-compat backends to save space
 ## Otherwise just run the normal build
@@ -364,14 +378,17 @@ COPY ./entrypoint.sh .
 
 # Copy the binary
 COPY --from=builder /build/local-ai ./
+# Copy the opus shim if it was built
+RUN --mount=from=builder,src=/build/,dst=/mnt/build \
+    if [ -f /mnt/build/libopusshim.so ]; then cp /mnt/build/libopusshim.so ./; fi
 
 # Make sure the models directory exists
-RUN mkdir -p /models /backends
+RUN mkdir -p /models /backends /data
 
 # Define the health check command
 HEALTHCHECK --interval=1m --timeout=10m --retries=10 \
   CMD curl -f ${HEALTHCHECK_ENDPOINT} || exit 1
 
-VOLUME /models /backends /configuration
+VOLUME /models /backends /configuration /data
 EXPOSE 8080
 ENTRYPOINT [ "/entrypoint.sh" ]

Dockerfile.aio

@@ -1,8 +0,0 @@
-ARG BASE_IMAGE=ubuntu:24.04
-
-FROM ${BASE_IMAGE} 
-
-RUN apt-get update && apt-get install -y pciutils && apt-get clean
-
-COPY aio/ /aio
-ENTRYPOINT [ "/aio/entrypoint.sh" ]

Makefile

@@ -1,5 +1,5 @@
 # Disable parallel execution for backend builds
-.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/outetts backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/stablediffusion-ggml-darwin backends/vllm backends/vllm-omni backends/moonshine backends/pocket-tts backends/qwen-tts backends/qwen-asr backends/nemo backends/voxcpm backends/whisperx backends/ace-step
+.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/outetts backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/mlx-distributed backends/stablediffusion-ggml-darwin backends/vllm backends/vllm-omni backends/moonshine backends/pocket-tts backends/qwen-tts backends/faster-qwen3-tts backends/qwen-asr backends/nemo backends/voxcpm backends/whisperx backends/ace-step backends/acestep-cpp backends/fish-speech backends/voxtral backends/opus backends/trl backends/llama-cpp-quantization
 
 GOCMD=go
 GOTEST=$(GOCMD) test
@@ -91,8 +91,23 @@ install-go-tools:
 	go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
 	go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
 
+## React UI:
+react-ui:
+ifneq ($(wildcard core/http/react-ui/dist),)
+	@echo "react-ui dist already exists, skipping build"
+else
+	cd core/http/react-ui && npm install && npm run build
+endif
+
+react-ui-docker:
+	docker run --entrypoint /bin/bash -v $(CURDIR):/app:z oven/bun:1 \
+	  -c "cd /app/core/http/react-ui && bun install && bun run build"
+
+core/http/react-ui/dist: react-ui
+
 ## Build:
-build: protogen-go install-go-tools ## Build the project
+
+build: protogen-go generate install-go-tools core/http/react-ui/dist ## Build the project
 	$(info ${GREEN}I local-ai build info:${RESET})
 	$(info ${GREEN}I BUILD_TYPE: ${YELLOW}$(BUILD_TYPE)${RESET})
 	$(info ${GREEN}I GO_TAGS: ${YELLOW}$(GO_TAGS)${RESET})
@@ -149,6 +164,7 @@ test: test-models/testmodel.ggml protogen-go
 	@echo 'Running tests'
 	export GO_TAGS="debug"
 	$(MAKE) prepare-test
+	OPUS_SHIM_LIBRARY=$(abspath ./pkg/opus/shim/libopusshim.so) \
 	HUGGINGFACE_GRPC=$(abspath ./)/backend/python/transformers/run.sh TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models BACKENDS_PATH=$(abspath ./)/backends \
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="!llama-gguf"  --flake-attempts $(TEST_FLAKES) --fail-fast -v -r $(TEST_PATHS)
 	$(MAKE) test-llama-gguf
@@ -156,10 +172,10 @@ test: test-models/testmodel.ggml protogen-go
 	$(MAKE) test-stablediffusion
 
 ########################################################
-## AIO tests
+## E2E AIO tests (uses standard image with pre-configured models)
 ########################################################
 
-docker-build-aio:
+docker-build-e2e:
 	docker build \
 		--build-arg MAKEFLAGS="--jobs=5 --output-sync=target" \
 		--build-arg BASE_IMAGE=$(BASE_IMAGE) \
@@ -171,13 +187,12 @@ docker-build-aio:
 		--build-arg UBUNTU_CODENAME=$(UBUNTU_CODENAME) \
 		--build-arg GO_TAGS="$(GO_TAGS)" \
 		-t local-ai:tests -f Dockerfile .
-	BASE_IMAGE=local-ai:tests DOCKER_AIO_IMAGE=local-ai-aio:test $(MAKE) docker-aio
 
 e2e-aio:
 	LOCALAI_BACKEND_DIR=$(abspath ./backends) \
-	LOCALAI_MODELS_DIR=$(abspath ./models) \
-	LOCALAI_IMAGE_TAG=test \
-	LOCALAI_IMAGE=local-ai-aio \
+	LOCALAI_MODELS_DIR=$(abspath ./tests/e2e-aio/models) \
+	LOCALAI_IMAGE_TAG=tests \
+	LOCALAI_IMAGE=local-ai \
 	$(MAKE) run-e2e-aio
 
 run-e2e-aio: protogen-go
@@ -236,6 +251,88 @@ test-stablediffusion: prepare-test
 test-stores:
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="stores" --flake-attempts $(TEST_FLAKES) -v -r tests/integration
 
+test-opus:
+	@echo 'Running opus backend tests'
+	$(MAKE) -C backend/go/opus libopusshim.so
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --flake-attempts $(TEST_FLAKES) -v -r ./backend/go/opus/...
+
+test-opus-docker:
+	@echo 'Running opus backend tests in Docker'
+	docker build --target builder \
+	  --build-arg BUILD_TYPE=$(or $(BUILD_TYPE),) \
+	  --build-arg BASE_IMAGE=$(or $(BASE_IMAGE),ubuntu:24.04) \
+	  --build-arg BACKEND=opus \
+	  -t localai-opus-test -f backend/Dockerfile.golang .
+	docker run --rm localai-opus-test \
+	  bash -c 'cd /LocalAI && go run github.com/onsi/ginkgo/v2/ginkgo --flake-attempts $(TEST_FLAKES) -v -r ./backend/go/opus/...'
+
+test-realtime: build-mock-backend
+	@echo 'Running realtime e2e tests (mock backend)'
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="Realtime && !real-models" --flake-attempts $(TEST_FLAKES) -v -r ./tests/e2e
+
+# Real-model realtime tests. Set REALTIME_TEST_MODEL to use your own pipeline,
+# or leave unset to auto-build one from the component env vars below.
+REALTIME_VAD?=silero-vad-ggml
+REALTIME_STT?=whisper-1
+REALTIME_LLM?=qwen3-0.6b
+REALTIME_TTS?=tts-1
+REALTIME_BACKENDS_PATH?=$(abspath ./)/backends
+
+test-realtime-models: build-mock-backend
+	@echo 'Running realtime e2e tests (real models)'
+	REALTIME_TEST_MODEL=$${REALTIME_TEST_MODEL:-realtime-test-pipeline} \
+	REALTIME_VAD=$(REALTIME_VAD) \
+	REALTIME_STT=$(REALTIME_STT) \
+	REALTIME_LLM=$(REALTIME_LLM) \
+	REALTIME_TTS=$(REALTIME_TTS) \
+	REALTIME_BACKENDS_PATH=$(REALTIME_BACKENDS_PATH) \
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="Realtime" --flake-attempts $(TEST_FLAKES) -v -r ./tests/e2e
+
+# --- Container-based real-model testing ---
+
+REALTIME_BACKEND_NAMES ?= silero-vad whisper llama-cpp kokoro
+REALTIME_MODELS_DIR ?= $(abspath ./models)
+REALTIME_BACKENDS_DIR ?= $(abspath ./local-backends)
+REALTIME_DOCKER_FLAGS ?= --gpus all
+
+local-backends:
+	mkdir -p local-backends
+
+extract-backend-%: docker-build-% local-backends
+	@echo "Extracting backend $*..."
+	@CID=$$(docker create local-ai-backend:$*) && \
+	  rm -rf local-backends/$* && mkdir -p local-backends/$* && \
+	  docker cp $$CID:/ - | tar -xf - -C local-backends/$* && \
+	  docker rm $$CID > /dev/null
+
+extract-realtime-backends: $(addprefix extract-backend-,$(REALTIME_BACKEND_NAMES))
+
+test-realtime-models-docker: build-mock-backend
+	docker build --target build-requirements \
+	  --build-arg BUILD_TYPE=$(or $(BUILD_TYPE),cublas) \
+	  --build-arg CUDA_MAJOR_VERSION=$(or $(CUDA_MAJOR_VERSION),13) \
+	  --build-arg CUDA_MINOR_VERSION=$(or $(CUDA_MINOR_VERSION),0) \
+	  -t localai-test-runner .
+	docker run --rm \
+	  $(REALTIME_DOCKER_FLAGS) \
+	  -v $(abspath ./):/build \
+	  -v $(REALTIME_MODELS_DIR):/models:ro \
+	  -v $(REALTIME_BACKENDS_DIR):/backends \
+	  -v localai-go-cache:/root/go/pkg/mod \
+	  -v localai-go-build-cache:/root/.cache/go-build \
+	  -e REALTIME_TEST_MODEL=$${REALTIME_TEST_MODEL:-realtime-test-pipeline} \
+	  -e REALTIME_VAD=$(REALTIME_VAD) \
+	  -e REALTIME_STT=$(REALTIME_STT) \
+	  -e REALTIME_LLM=$(REALTIME_LLM) \
+	  -e REALTIME_TTS=$(REALTIME_TTS) \
+	  -e REALTIME_BACKENDS_PATH=/backends \
+	  -e REALTIME_MODELS_PATH=/models \
+	  -w /build \
+	  localai-test-runner \
+	  bash -c 'git config --global --add safe.directory /build && \
+	    make protogen-go && make build-mock-backend && \
+	    go run github.com/onsi/ginkgo/v2/ginkgo --label-filter="Realtime" --flake-attempts $(TEST_FLAKES) -v -r ./tests/e2e'
+
 test-container:
 	docker build --target requirements -t local-ai-test-container .
 	docker run -ti --rm --entrypoint /bin/bash -ti -v $(abspath ./):/build local-ai-test-container
@@ -301,6 +398,16 @@ protogen-go: protoc install-go-tools
 	./protoc --experimental_allow_proto3_optional -Ibackend/ --go_out=pkg/grpc/proto/ --go_opt=paths=source_relative --go-grpc_out=pkg/grpc/proto/ --go-grpc_opt=paths=source_relative \
     backend/backend.proto
 
+core/config/inference_defaults.json: ## Fetch inference defaults from unsloth (only if missing)
+	$(GOCMD) generate ./core/config/...
+
+.PHONY: generate
+generate: core/config/inference_defaults.json ## Ensure inference defaults exist
+
+.PHONY: generate-force
+generate-force: ## Re-fetch inference defaults from unsloth (always)
+	$(GOCMD) generate ./core/config/...
+
 .PHONY: protogen-go-clean
 protogen-go-clean:
 	$(RM) pkg/grpc/proto/backend.pb.go pkg/grpc/proto/backend_grpc.pb.go
@@ -317,11 +424,14 @@ prepare-test-extra: protogen-python
 	$(MAKE) -C backend/python/moonshine
 	$(MAKE) -C backend/python/pocket-tts
 	$(MAKE) -C backend/python/qwen-tts
+	$(MAKE) -C backend/python/fish-speech
+	$(MAKE) -C backend/python/faster-qwen3-tts
 	$(MAKE) -C backend/python/qwen-asr
 	$(MAKE) -C backend/python/nemo
 	$(MAKE) -C backend/python/voxcpm
 	$(MAKE) -C backend/python/whisperx
 	$(MAKE) -C backend/python/ace-step
+	$(MAKE) -C backend/python/trl
 
 test-extra: prepare-test-extra
 	$(MAKE) -C backend/python/transformers test
@@ -334,14 +444,16 @@ test-extra: prepare-test-extra
 	$(MAKE) -C backend/python/moonshine test
 	$(MAKE) -C backend/python/pocket-tts test
 	$(MAKE) -C backend/python/qwen-tts test
+	$(MAKE) -C backend/python/fish-speech test
+	$(MAKE) -C backend/python/faster-qwen3-tts test
 	$(MAKE) -C backend/python/qwen-asr test
 	$(MAKE) -C backend/python/nemo test
 	$(MAKE) -C backend/python/voxcpm test
 	$(MAKE) -C backend/python/whisperx test
 	$(MAKE) -C backend/python/ace-step test
+	$(MAKE) -C backend/python/trl test
 
 DOCKER_IMAGE?=local-ai
-DOCKER_AIO_IMAGE?=local-ai-aio
 IMAGE_TYPE?=core
 BASE_IMAGE?=ubuntu:24.04
 
@@ -371,21 +483,6 @@ docker-cuda12:
 		--build-arg UBUNTU_CODENAME=$(UBUNTU_CODENAME) \
 		-t $(DOCKER_IMAGE)-cuda-12 .
 
-docker-aio:
-	@echo "Building AIO image with base $(BASE_IMAGE) as $(DOCKER_AIO_IMAGE)"
-	docker build \
-		--build-arg BASE_IMAGE=$(BASE_IMAGE) \
-		--build-arg MAKEFLAGS="$(DOCKER_MAKEFLAGS)" \
-		--build-arg CUDA_MAJOR_VERSION=$(CUDA_MAJOR_VERSION) \
-		--build-arg CUDA_MINOR_VERSION=$(CUDA_MINOR_VERSION) \
-		--build-arg UBUNTU_VERSION=$(UBUNTU_VERSION) \
-		--build-arg UBUNTU_CODENAME=$(UBUNTU_CODENAME) \
-		-t $(DOCKER_AIO_IMAGE) -f Dockerfile.aio .
-
-docker-aio-all:
-	$(MAKE) docker-aio DOCKER_AIO_SIZE=cpu
-	$(MAKE) docker-aio DOCKER_AIO_SIZE=cpu
-
 docker-image-intel:
 	docker build \
 		--build-arg BASE_IMAGE=intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04 \
@@ -435,6 +532,10 @@ backends/mlx-audio:
 	BACKEND=mlx-audio $(MAKE) build-darwin-python-backend
 	./local-ai backends install "ocifile://$(abspath ./backend-images/mlx-audio.tar)"
 
+backends/mlx-distributed:
+	BACKEND=mlx-distributed $(MAKE) build-darwin-python-backend
+	./local-ai backends install "ocifile://$(abspath ./backend-images/mlx-distributed.tar)"
+
 backends/stablediffusion-ggml-darwin:
 	BACKEND=stablediffusion-ggml BUILD_TYPE=metal $(MAKE) build-darwin-go-backend
 	./local-ai backends install "ocifile://$(abspath ./backend-images/stablediffusion-ggml.tar)"
@@ -443,8 +544,9 @@ backend-images:
 	mkdir -p backend-images
 
 # Backend metadata: BACKEND_NAME | DOCKERFILE_TYPE | BUILD_CONTEXT | PROGRESS_FLAG | NEEDS_BACKEND_ARG
-# llama-cpp is special - uses llama-cpp Dockerfile and doesn't need BACKEND arg
+# llama-cpp and forks - use llama-cpp Dockerfile
 BACKEND_LLAMA_CPP = llama-cpp|llama-cpp|.|false|false
+BACKEND_LLAMA_CPP_TQ = llama-cpp-tq|llama-cpp|.|false|true
 
 # Golang backends
 BACKEND_PIPER = piper|golang|.|false|true
@@ -453,6 +555,9 @@ BACKEND_HUGGINGFACE = huggingface|golang|.|false|true
 BACKEND_SILERO_VAD = silero-vad|golang|.|false|true
 BACKEND_STABLEDIFFUSION_GGML = stablediffusion-ggml|golang|.|--progress=plain|true
 BACKEND_WHISPER = whisper|golang|.|false|true
+BACKEND_VOXTRAL = voxtral|golang|.|false|true
+BACKEND_ACESTEP_CPP = acestep-cpp|golang|.|false|true
+BACKEND_OPUS = opus|golang|.|false|true
 
 # Python backends with root context
 BACKEND_RERANKERS = rerankers|python|.|false|true
@@ -472,11 +577,16 @@ BACKEND_VIBEVOICE = vibevoice|python|.|--progress=plain|true
 BACKEND_MOONSHINE = moonshine|python|.|false|true
 BACKEND_POCKET_TTS = pocket-tts|python|.|false|true
 BACKEND_QWEN_TTS = qwen-tts|python|.|false|true
+BACKEND_FISH_SPEECH = fish-speech|python|.|false|true
+BACKEND_FASTER_QWEN3_TTS = faster-qwen3-tts|python|.|false|true
 BACKEND_QWEN_ASR = qwen-asr|python|.|false|true
 BACKEND_NEMO = nemo|python|.|false|true
 BACKEND_VOXCPM = voxcpm|python|.|false|true
 BACKEND_WHISPERX = whisperx|python|.|false|true
 BACKEND_ACE_STEP = ace-step|python|.|false|true
+BACKEND_MLX_DISTRIBUTED = mlx-distributed|python|./|false|true
+BACKEND_TRL = trl|python|.|false|true
+BACKEND_LLAMA_CPP_QUANTIZATION = llama-cpp-quantization|python|.|false|true
 
 # Helper function to build docker image for a backend
 # Usage: $(call docker-build-backend,BACKEND_NAME,DOCKERFILE_TYPE,BUILD_CONTEXT,PROGRESS_FLAG,NEEDS_BACKEND_ARG)
@@ -500,12 +610,15 @@ endef
 
 # Generate all docker-build targets
 $(eval $(call generate-docker-build-target,$(BACKEND_LLAMA_CPP)))
+$(eval $(call generate-docker-build-target,$(BACKEND_LLAMA_CPP_TQ)))
 $(eval $(call generate-docker-build-target,$(BACKEND_PIPER)))
 $(eval $(call generate-docker-build-target,$(BACKEND_LOCAL_STORE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_HUGGINGFACE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_SILERO_VAD)))
 $(eval $(call generate-docker-build-target,$(BACKEND_STABLEDIFFUSION_GGML)))
 $(eval $(call generate-docker-build-target,$(BACKEND_WHISPER)))
+$(eval $(call generate-docker-build-target,$(BACKEND_VOXTRAL)))
+$(eval $(call generate-docker-build-target,$(BACKEND_OPUS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_RERANKERS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_TRANSFORMERS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_OUTETTS)))
@@ -523,17 +636,23 @@ $(eval $(call generate-docker-build-target,$(BACKEND_VIBEVOICE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_MOONSHINE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_POCKET_TTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_QWEN_TTS)))
+$(eval $(call generate-docker-build-target,$(BACKEND_FISH_SPEECH)))
+$(eval $(call generate-docker-build-target,$(BACKEND_FASTER_QWEN3_TTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_QWEN_ASR)))
 $(eval $(call generate-docker-build-target,$(BACKEND_NEMO)))
 $(eval $(call generate-docker-build-target,$(BACKEND_VOXCPM)))
 $(eval $(call generate-docker-build-target,$(BACKEND_WHISPERX)))
 $(eval $(call generate-docker-build-target,$(BACKEND_ACE_STEP)))
+$(eval $(call generate-docker-build-target,$(BACKEND_ACESTEP_CPP)))
+$(eval $(call generate-docker-build-target,$(BACKEND_MLX_DISTRIBUTED)))
+$(eval $(call generate-docker-build-target,$(BACKEND_TRL)))
+$(eval $(call generate-docker-build-target,$(BACKEND_LLAMA_CPP_QUANTIZATION)))
 
 # Pattern rule for docker-save targets
 docker-save-%: backend-images
 	docker save local-ai-backend:$* -o backend-images/$*.tar
 
-docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-vllm-omni docker-build-transformers docker-build-outetts docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-chatterbox docker-build-vibevoice docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts docker-build-qwen-asr docker-build-nemo docker-build-voxcpm docker-build-whisperx docker-build-ace-step
+docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-vllm-omni docker-build-transformers docker-build-outetts docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-chatterbox docker-build-vibevoice docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts docker-build-fish-speech docker-build-faster-qwen3-tts docker-build-qwen-asr docker-build-nemo docker-build-voxcpm docker-build-whisperx docker-build-ace-step docker-build-acestep-cpp docker-build-voxtral docker-build-mlx-distributed docker-build-trl docker-build-llama-cpp-quantization
 
 ########################################################
 ### Mock Backend for E2E Tests
@@ -545,6 +664,23 @@ build-mock-backend: protogen-go
 clean-mock-backend:
 	rm -f tests/e2e/mock-backend/mock-backend
 
+########################################################
+### UI E2E Test Server
+########################################################
+
+build-ui-test-server: build-mock-backend react-ui protogen-go
+	$(GOCMD) build -o tests/e2e-ui/ui-test-server ./tests/e2e-ui
+
+test-ui-e2e: build-ui-test-server
+	cd core/http/react-ui && npm install && npx playwright install --with-deps chromium && npx playwright test
+
+test-ui-e2e-docker:
+	docker build -t localai-ui-e2e -f tests/e2e-ui/Dockerfile .
+	docker run --rm localai-ui-e2e
+
+clean-ui-test-server:
+	rm -f tests/e2e-ui/ui-test-server
+
 ########################################################
 ### END Backends
 ########################################################
@@ -553,6 +689,7 @@ clean-mock-backend:
 swagger:
 	swag init -g core/http/app.go --output swagger
 
+# DEPRECATED: gen-assets is for the legacy Alpine.js UI. Remove when legacy UI is removed.
 .PHONY: gen-assets
 gen-assets:
 	$(GOCMD) run core/dependencies_manager/manager.go webui_static.yaml core/http/static/assets

README.md

@@ -5,26 +5,14 @@
 </h1>
 
 <p align="center">
-<a href="https://github.com/go-skynet/LocalAI/fork" target="blank">
-<img src="https://img.shields.io/github/forks/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI forks"/>
-</a>
 <a href="https://github.com/go-skynet/LocalAI/stargazers" target="blank">
 <img src="https://img.shields.io/github/stars/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI stars"/>
 </a>
-<a href="https://github.com/go-skynet/LocalAI/pulls" target="blank">
-<img src="https://img.shields.io/github/issues-pr/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI pull-requests"/>
-</a>
 <a href='https://github.com/go-skynet/LocalAI/releases'>
 <img src='https://img.shields.io/github/release/go-skynet/LocalAI?&label=Latest&style=for-the-badge'>
 </a>
-</p>
-
-<p align="center">
-<a href="https://hub.docker.com/r/localai/localai" target="blank">
-<img src="https://img.shields.io/badge/dockerhub-images-important.svg?logo=Docker" alt="LocalAI Docker hub"/>
-</a>
-<a href="https://quay.io/repository/go-skynet/local-ai?tab=tags&tag=latest" target="blank">
-<img src="https://img.shields.io/badge/quay.io-images-important.svg?" alt="LocalAI Quay.io"/>
+<a href="LICENSE" target="blank">
+<img src="https://img.shields.io/badge/License-MIT-yellow.svg?style=for-the-badge" alt="LocalAI License"/>
 </a>
 </p>
 
@@ -41,338 +29,183 @@
 <a href="https://trendshift.io/repositories/5539" target="_blank"><img src="https://trendshift.io/api/badge/repositories/5539" alt="mudler%2FLocalAI | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
 </p>
 
-> :bulb: Get help - [❓FAQ](https://localai.io/faq/) [💭Discussions](https://github.com/go-skynet/LocalAI/discussions) [:speech_balloon: Discord](https://discord.gg/uJAeKSAGDy) [:book: Documentation website](https://localai.io/)
->
-> [💻 Quickstart](https://localai.io/basics/getting_started/) [🖼️ Models](https://models.localai.io/) [🚀 Roadmap](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap) [🛫 Examples](https://github.com/mudler/LocalAI-examples) Try on 
-[![Telegram](https://img.shields.io/badge/Telegram-2CA5E0?style=for-the-badge&logo=telegram&logoColor=white)](https://t.me/localaiofficial_bot)
+**LocalAI** is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
 
-[![tests](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml)[![Build and Release](https://github.com/go-skynet/LocalAI/actions/workflows/release.yaml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/release.yaml)[![build container images](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml)[![Bump dependencies](https://github.com/go-skynet/LocalAI/actions/workflows/bump_deps.yaml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/bump_deps.yaml)[![Artifact Hub](https://img.shields.io/endpoint?url=https://artifacthub.io/badge/repository/localai)](https://artifacthub.io/packages/search?repo=localai)
+- **Drop-in API compatibility** — OpenAI, Anthropic, ElevenLabs APIs
+- **35+ backends** — llama.cpp, vLLM, transformers, whisper, diffusers, MLX...
+- **Any hardware** — NVIDIA, AMD, Intel, Apple Silicon, Vulkan, or CPU-only
+- **Multi-user ready** — API key auth, user quotas, role-based access
+- **Built-in AI agents** — autonomous agents with tool use, RAG, MCP, and skills
+- **Privacy-first** — your data never leaves your infrastructure
 
-**LocalAI** is the free, Open Source OpenAI alternative. LocalAI act as a drop-in replacement REST API that's compatible with OpenAI (Elevenlabs, Anthropic... ) API specifications for local AI inferencing. It allows you to run LLMs, generate images, audio (and not only) locally or on-prem with consumer grade hardware, supporting multiple model families. Does not require GPU. It is created and maintained by [Ettore Di Giacinto](https://github.com/mudler).
+Created and maintained by [Ettore Di Giacinto](https://github.com/mudler).
 
+> [:book: Documentation](https://localai.io/) | [:speech_balloon: Discord](https://discord.gg/uJAeKSAGDy) | [💻 Quickstart](https://localai.io/basics/getting_started/) | [🖼️ Models](https://models.localai.io/) | [❓FAQ](https://localai.io/faq/)
 
-## Local Stack Family
+## Guided tour
 
-Liking LocalAI? LocalAI is part of an integrated suite of AI infrastructure tools, you might also like:
+https://github.com/user-attachments/assets/08cbb692-57da-48f7-963d-2e7b43883c18
 
-- **[LocalAGI](https://github.com/mudler/LocalAGI)** - AI agent orchestration platform with OpenAI Responses API compatibility and advanced agentic capabilities
-- **[LocalRecall](https://github.com/mudler/LocalRecall)** - MCP/REST API knowledge base system providing persistent memory and storage for AI agents
-- 🆕 **[Cogito](https://github.com/mudler/cogito)** - Go library for building intelligent, co-operative agentic software and LLM-powered workflows, focusing on improving results for small, open source language models that scales to any LLM. Powers LocalAGI and LocalAI MCP/Agentic capabilities
-- 🆕 **[Wiz](https://github.com/mudler/wiz)** - Terminal-based AI agent accessible via Ctrl+Space keybinding. Portable, local-LLM friendly shell assistant with TUI/CLI modes, tool execution with approval, MCP protocol support, and multi-shell compatibility (zsh, bash, fish)
-- 🆕 **[SkillServer](https://github.com/mudler/skillserver)** - Simple, centralized skills database for AI agents via MCP. Manages skills as Markdown files with MCP server integration, web UI for editing, Git synchronization, and full-text search capabilities
+<details>
 
+<summary>
+Click to see more!
+</summary>
 
-## Screenshots / Video
+#### User and auth
 
-### Youtube video
+https://github.com/user-attachments/assets/228fa9ad-81a3-4d43-bfb9-31557e14a36c
 
-<h1 align="center">
-  <br>
-  <a href="https://www.youtube.com/watch?v=PDqYhB9nNHA" target="_blank"> <img width="300" src="https://img.youtube.com/vi/PDqYhB9nNHA/0.jpg"> </a><br>
-<br>
-</h1>
+#### Agents
 
+https://github.com/user-attachments/assets/6270b331-e21d-4087-a540-6290006b381a
 
-### Screenshots
+#### Usage metrics per user
 
-| Talk Interface | Generate Audio |
-| --- | --- |
-| ![Screenshot 2025-03-31 at 12-01-36 LocalAI - Talk](./docs/assets/images/screenshots/screenshot_tts.png) | ![Screenshot 2025-03-31 at 12-01-29 LocalAI - Generate audio with voice-en-us-ryan-low](./docs/assets/images/screenshots/screenshot_tts.png) |
+https://github.com/user-attachments/assets/cbb03379-23b4-4e3d-bd26-d152f057007f
 
-| Models Overview | Generate Images |
-| --- | --- |
-| ![Screenshot 2025-03-31 at 12-01-20 LocalAI - Models](./docs/assets/images/screenshots/screenshot_gallery.png) | ![Screenshot 2025-03-31 at 12-31-41 LocalAI - Generate images with flux 1-dev](./docs/assets/images/screenshots/screenshot_image.png) |
+#### Fine-tuning and Quantization
 
-| Chat Interface | Home |
-| --- | --- |
-| ![Screenshot 2025-03-31 at 11-57-44 LocalAI - Chat with localai-functioncall-qwen2 5-7b-v0 5](./docs/assets/images/screenshots/screenshot_chat.png) | ![Screenshot 2025-03-31 at 11-57-23 LocalAI API - c2a39e3 (c2a39e3639227cfd94ffffe9f5691239acc275a8)](./docs/assets/images/screenshots/screenshot_home.png) |
+https://github.com/user-attachments/assets/5ba4ace9-d3df-4795-b7d4-b0b404ea71ee
 
-| Login | Swarm |
-| --- | --- |
-|![Screenshot 2025-03-31 at 12-09-59 ](./docs/assets/images/screenshots/screenshot_login.png) | ![Screenshot 2025-03-31 at 12-10-39 LocalAI - P2P dashboard](./docs/assets/images/screenshots/screenshot_p2p.png) |
+#### WebRTC
 
-## 💻 Quickstart
+https://github.com/user-attachments/assets/ed88e34c-fed3-4b83-8a67-4716a9feeb7b
 
-> ⚠️ **Note:** The `install.sh` script is currently experiencing issues due to the heavy changes currently undergoing in LocalAI and may produce broken or misconfigured installations. Please use Docker installation (see below) or manual binary installation until [issue #8032](https://github.com/mudler/LocalAI/issues/8032) is resolved.
+</details>
 
-Run the installer script:
+## Quickstart
 
-```bash
-# Basic installation
-curl https://localai.io/install.sh | sh
-```
-
-For more installation options, see [Installer Options](https://localai.io/installation/).
-
-### macOS Download:
+### macOS
 
 <a href="https://github.com/mudler/LocalAI/releases/latest/download/LocalAI.dmg">
   <img src="https://img.shields.io/badge/Download-macOS-blue?style=for-the-badge&logo=apple&logoColor=white" alt="Download LocalAI for macOS"/>
 </a>
 
-> Note: the DMGs are not signed by Apple as quarantined. See https://github.com/mudler/LocalAI/issues/6268 for a workaround, fix is tracked here: https://github.com/mudler/LocalAI/issues/6244
+> **Note:** The DMG is not signed by Apple. After installing, run: `sudo xattr -d com.apple.quarantine /Applications/LocalAI.app`. See [#6268](https://github.com/mudler/LocalAI/issues/6268) for details.
 
 ### Containers (Docker, podman, ...)
 
-> **💡 Docker Run vs Docker Start**
-> 
-> - `docker run` creates and starts a new container. If a container with the same name already exists, this command will fail.
-> - `docker start` starts an existing container that was previously created with `docker run`.
-> 
-> If you've already run LocalAI before and want to start it again, use: `docker start -i local-ai`
+> Already ran LocalAI before? Use `docker start -i local-ai` to restart an existing container.
 
-#### CPU only image:
+#### CPU only:
 
 ```bash
 docker run -ti --name local-ai -p 8080:8080 localai/localai:latest
 ```
 
-#### NVIDIA GPU Images:
+#### NVIDIA GPU:
 
 ```bash
-# CUDA 13.0
+# CUDA 13
 docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-gpu-nvidia-cuda-13
 
-# CUDA 12.0
+# CUDA 12
 docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-gpu-nvidia-cuda-12
 
-# NVIDIA Jetson (L4T) ARM64
-# CUDA 12 (for Nvidia AGX Orin and similar platforms)
+# NVIDIA Jetson ARM64 (CUDA 12, for AGX Orin and similar)
 docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-nvidia-l4t-arm64
 
-# CUDA 13 (for Nvidia DGX Spark)
+# NVIDIA Jetson ARM64 (CUDA 13, for DGX Spark)
 docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-nvidia-l4t-arm64-cuda-13
 ```
 
-#### AMD GPU Images (ROCm):
+#### AMD GPU (ROCm):
 
 ```bash
 docker run -ti --name local-ai -p 8080:8080 --device=/dev/kfd --device=/dev/dri --group-add=video localai/localai:latest-gpu-hipblas
 ```
 
-#### Intel GPU Images (oneAPI):
+#### Intel GPU (oneAPI):
 
 ```bash
 docker run -ti --name local-ai -p 8080:8080 --device=/dev/dri/card1 --device=/dev/dri/renderD128 localai/localai:latest-gpu-intel
 ```
 
-#### Vulkan GPU Images:
+#### Vulkan GPU:
 
 ```bash
 docker run -ti --name local-ai -p 8080:8080 localai/localai:latest-gpu-vulkan
 ```
 
-#### AIO Images (pre-downloaded models):
+### Loading models
 
 ```bash
-# CPU version
-docker run -ti --name local-ai -p 8080:8080 localai/localai:latest-aio-cpu
-
-# NVIDIA CUDA 13 version
-docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-aio-gpu-nvidia-cuda-13
-
-# NVIDIA CUDA 12 version
-docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-aio-gpu-nvidia-cuda-12
-
-# Intel GPU version
-docker run -ti --name local-ai -p 8080:8080 localai/localai:latest-aio-gpu-intel
-
-# AMD GPU version
-docker run -ti --name local-ai -p 8080:8080 --device=/dev/kfd --device=/dev/dri --group-add=video localai/localai:latest-aio-gpu-hipblas
-```
-
-For more information about the AIO images and pre-downloaded models, see [Container Documentation](https://localai.io/basics/container/).
-
-To load models:
-
-```bash
-# From the model gallery (see available models with `local-ai models list`, in the WebUI from the model tab, or visiting https://models.localai.io)
+# From the model gallery (see available models with `local-ai models list` or at https://models.localai.io)
 local-ai run llama-3.2-1b-instruct:q4_k_m
-# Start LocalAI with the phi-2 model directly from huggingface
+# From Huggingface
 local-ai run huggingface://TheBloke/phi-2-GGUF/phi-2.Q8_0.gguf
-# Install and run a model from the Ollama OCI registry
+# From the Ollama OCI registry
 local-ai run ollama://gemma:2b
-# Run a model from a configuration file
+# From a YAML config
 local-ai run https://gist.githubusercontent.com/.../phi-2.yaml
-# Install and run a model from a standard OCI registry (e.g., Docker Hub)
+# From a standard OCI registry (e.g., Docker Hub)
 local-ai run oci://localai/phi-2:latest
 ```
 
-> ⚡ **Automatic Backend Detection**: When you install models from the gallery or YAML files, LocalAI automatically detects your system's GPU capabilities (NVIDIA, AMD, Intel) and downloads the appropriate backend. For advanced configuration options, see [GPU Acceleration](https://localai.io/features/gpu-acceleration/#automatic-backend-detection).
+> **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 information, see [💻 Getting started](https://localai.io/basics/getting_started/index.html), if you are interested in our roadmap items and future enhancements, you can see the [Issues labeled as Roadmap here](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
+For more details, see the [Getting Started guide](https://localai.io/basics/getting_started/).
 
-## 📰 Latest project news
-- February 2026: [Realtime API for audio-to-audio with tool calling](https://github.com/mudler/LocalAI/pull/6245), [ACE-Step 1.5 support](https://github.com/mudler/LocalAI/pull/8396)
-- January 2026: **LocalAI 3.10.0** - Major release with Anthropic API support, Open Responses API for stateful agents, video & image generation suite (LTX-2), unified GPU backends, tool streaming & XML parsing, system-aware backend gallery, crash fixes for AVX-only CPUs and AMD VRAM reporting, request tracing, and new backends: **Moonshine** (ultra-fast transcription), **Pocket-TTS** (lightweight TTS). Vulkan arm64 builds now available. [Release notes](https://github.com/mudler/LocalAI/releases/tag/v3.10.0).
-- December 2025: [Dynamic Memory Resource reclaimer](https://github.com/mudler/LocalAI/pull/7583), [Automatic fitting of models to multiple GPUS(llama.cpp)](https://github.com/mudler/LocalAI/pull/7584), [Added Vibevoice backend](https://github.com/mudler/LocalAI/pull/7494)
-- November 2025: Major improvements to the UX. Among these: [Import models via URL](https://github.com/mudler/LocalAI/pull/7245) and [Multiple chats and history](https://github.com/mudler/LocalAI/pull/7325)
-- October 2025: 🔌 [Model Context Protocol (MCP)](https://localai.io/docs/features/mcp/) support added for agentic capabilities with external tools
-- September 2025: New Launcher application for MacOS and Linux, extended support to many backends for Mac and Nvidia L4T devices. Models: Added MLX-Audio, WAN 2.2. WebUI improvements and Python-based backends now ships portable python environments.
-- August 2025: MLX, MLX-VLM, Diffusers and llama.cpp are now supported on Mac M1/M2/M3+ chips ( with `development` suffix in the gallery ): https://github.com/mudler/LocalAI/pull/6049 https://github.com/mudler/LocalAI/pull/6119 https://github.com/mudler/LocalAI/pull/6121 https://github.com/mudler/LocalAI/pull/6060
-- July/August 2025: 🔍 [Object Detection](https://localai.io/features/object-detection/) added to the API featuring [rf-detr](https://github.com/roboflow/rf-detr)
-- July 2025: All backends migrated outside of the main binary. LocalAI is now more lightweight, small, and automatically downloads the required backend to run the model. [Read the release notes](https://github.com/mudler/LocalAI/releases/tag/v3.2.0)
-- June 2025: [Backend management](https://github.com/mudler/LocalAI/pull/5607) has been added. Attention: extras images are going to be deprecated from the next release! Read [the backend management PR](https://github.com/mudler/LocalAI/pull/5607).
-- May 2025: [Audio input](https://github.com/mudler/LocalAI/pull/5466) and [Reranking](https://github.com/mudler/LocalAI/pull/5396) in llama.cpp backend, [Realtime API](https://github.com/mudler/LocalAI/pull/5392),  Support to Gemma, SmollVLM, and more multimodal models (available in the gallery).
-- May 2025: Important: image name changes [See release](https://github.com/mudler/LocalAI/releases/tag/v2.29.0)
-- Apr 2025: Rebrand, WebUI enhancements
-- Apr 2025: [LocalAGI](https://github.com/mudler/LocalAGI) and [LocalRecall](https://github.com/mudler/LocalRecall) join the LocalAI family stack.
-- Apr 2025: WebUI overhaul, AIO images updates
-- Feb 2025: Backend cleanup, Breaking changes, new backends (kokoro, OutelTTS, faster-whisper), Nvidia L4T images
-- Jan 2025: LocalAI model release: https://huggingface.co/mudler/LocalAI-functioncall-phi-4-v0.3, SANA support in diffusers: https://github.com/mudler/LocalAI/pull/4603
-- Dec 2024: stablediffusion.cpp backend (ggml) added ( https://github.com/mudler/LocalAI/pull/4289 )
-- Nov 2024: Bark.cpp backend added ( https://github.com/mudler/LocalAI/pull/4287 )
-- Nov 2024: Voice activity detection models (**VAD**) added to the API: https://github.com/mudler/LocalAI/pull/4204
-- Oct 2024: examples moved to [LocalAI-examples](https://github.com/mudler/LocalAI-examples)
-- Aug 2024:  🆕 FLUX-1, [P2P Explorer](https://explorer.localai.io)
-- July 2024: 🔥🔥 🆕 P2P Dashboard, LocalAI Federated mode and AI Swarms: https://github.com/mudler/LocalAI/pull/2723. P2P Global community pools: https://github.com/mudler/LocalAI/issues/3113
-- May 2024: 🔥🔥 Decentralized P2P llama.cpp:  https://github.com/mudler/LocalAI/pull/2343 (peer2peer llama.cpp!) 👉 Docs  https://localai.io/features/distribute/
-- May 2024: 🔥🔥 Distributed inferencing: https://github.com/mudler/LocalAI/pull/2324
-- April 2024: Reranker API: https://github.com/mudler/LocalAI/pull/2121
+## Latest News
 
-Roadmap items: [List of issues](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
+- **March 2026**: [Agent management](https://github.com/mudler/LocalAI/pull/8820), [New React UI](https://github.com/mudler/LocalAI/pull/8772), [WebRTC](https://github.com/mudler/LocalAI/pull/8790), [MLX-distributed via P2P and RDMA](https://github.com/mudler/LocalAI/pull/8801), [MCP Apps, MCP Client-side](https://github.com/mudler/LocalAI/pull/8947)
+- **February 2026**: [Realtime API for audio-to-audio with tool calling](https://github.com/mudler/LocalAI/pull/6245), [ACE-Step 1.5 support](https://github.com/mudler/LocalAI/pull/8396)
+- **January 2026**: **LocalAI 3.10.0** — Anthropic API support, Open Responses API, video & image generation (LTX-2), unified GPU backends, tool streaming, Moonshine, Pocket-TTS. [Release notes](https://github.com/mudler/LocalAI/releases/tag/v3.10.0)
+- **December 2025**: [Dynamic Memory Resource reclaimer](https://github.com/mudler/LocalAI/pull/7583), [Automatic multi-GPU model fitting (llama.cpp)](https://github.com/mudler/LocalAI/pull/7584), [Vibevoice backend](https://github.com/mudler/LocalAI/pull/7494)
+- **November 2025**: [Import models via URL](https://github.com/mudler/LocalAI/pull/7245), [Multiple chats and history](https://github.com/mudler/LocalAI/pull/7325)
+- **October 2025**: [Model Context Protocol (MCP)](https://localai.io/docs/features/mcp/) support for agentic capabilities
+- **September 2025**: New Launcher for macOS and Linux, extended backend support for Mac and Nvidia L4T, MLX-Audio, WAN 2.2
+- **August 2025**: MLX, MLX-VLM, Diffusers, llama.cpp now supported on Apple Silicon
+- **July 2025**: All backends migrated outside the main binary — [lightweight, modular architecture](https://github.com/mudler/LocalAI/releases/tag/v3.2.0)
 
-## 🚀 [Features](https://localai.io/features/)
+For older news and full release notes, see [GitHub Releases](https://github.com/mudler/LocalAI/releases) and the [News page](https://localai.io/basics/news/).
 
-- 🧩 [Backend Gallery](https://localai.io/backends/): Install/remove backends on the fly, powered by OCI images — fully customizable and API-driven.
-- 📖 [Text generation with GPTs](https://localai.io/features/text-generation/) (`llama.cpp`, `transformers`, `vllm` ... [:book: and more](https://localai.io/model-compatibility/index.html#model-compatibility-table))
-- 🗣 [Text to Audio](https://localai.io/features/text-to-audio/)
-- 🔈 [Audio to Text](https://localai.io/features/audio-to-text/) (Audio transcription with `whisper.cpp`)
-- 🎨 [Image generation](https://localai.io/features/image-generation)
-- 🔥 [OpenAI-alike tools API](https://localai.io/features/openai-functions/) 
-- ⚡ [Realtime API](https://localai.io/features/openai-realtime/) (Speech-to-speech) 
-- 🧠 [Embeddings generation for vector databases](https://localai.io/features/embeddings/)
-- ✍️ [Constrained grammars](https://localai.io/features/constrained_grammars/)
-- 🖼️ [Download Models directly from Huggingface ](https://localai.io/models/)
-- 🥽 [Vision API](https://localai.io/features/gpt-vision/)
-- 🔍 [Object Detection](https://localai.io/features/object-detection/)
-- 📈 [Reranker API](https://localai.io/features/reranker/)
-- 🆕🖧 [P2P Inferencing](https://localai.io/features/distribute/)
-- 🆕🔌 [Model Context Protocol (MCP)](https://localai.io/docs/features/mcp/) - Agentic capabilities with external tools and [LocalAGI's Agentic capabilities](https://github.com/mudler/LocalAGI)
-- 🔊 Voice activity detection (Silero-VAD support)
-- 🌍 Integrated WebUI!
+## Features
 
-## 🧩 Supported Backends & Acceleration
+- [Text generation](https://localai.io/features/text-generation/) (`llama.cpp`, `transformers`, `vllm` ... [and more](https://localai.io/model-compatibility/))
+- [Text to Audio](https://localai.io/features/text-to-audio/)
+- [Audio to Text](https://localai.io/features/audio-to-text/)
+- [Image generation](https://localai.io/features/image-generation)
+- [OpenAI-compatible tools API](https://localai.io/features/openai-functions/)
+- [Realtime API](https://localai.io/features/openai-realtime/) (Speech-to-speech)
+- [Embeddings generation](https://localai.io/features/embeddings/)
+- [Constrained grammars](https://localai.io/features/constrained_grammars/)
+- [Download models from Huggingface](https://localai.io/models/)
+- [Vision API](https://localai.io/features/gpt-vision/)
+- [Object Detection](https://localai.io/features/object-detection/)
+- [Reranker API](https://localai.io/features/reranker/)
+- [P2P Inferencing](https://localai.io/features/distribute/)
+- [Distributed Mode](https://localai.io/features/distributed-mode/) — Horizontal scaling with PostgreSQL + NATS
+- [Model Context Protocol (MCP)](https://localai.io/docs/features/mcp/)
+- [Built-in Agents](https://localai.io/features/agents/) — Autonomous AI agents with tool use, RAG, skills, SSE streaming, and [Agent Hub](https://agenthub.localai.io)
+- [Backend Gallery](https://localai.io/backends/) — Install/remove backends on the fly via OCI images
+- Voice Activity Detection (Silero-VAD)
+- Integrated WebUI
 
-LocalAI supports a comprehensive range of AI backends with multiple acceleration options:
+## Supported Backends & Acceleration
 
-### Text Generation & Language Models
-| Backend | Description | Acceleration Support |
-|---------|-------------|---------------------|
-| **llama.cpp** | LLM inference in C/C++ | CUDA 12/13, ROCm, Intel SYCL, Vulkan, Metal, CPU |
-| **vLLM** | Fast LLM inference with PagedAttention | CUDA 12/13, ROCm, Intel |
-| **transformers** | HuggingFace transformers framework | CUDA 12/13, ROCm, Intel, CPU |
-| **MLX** | Apple Silicon LLM inference | Metal (M1/M2/M3+) |
-| **MLX-VLM** | Apple Silicon Vision-Language Models | Metal (M1/M2/M3+) |
+LocalAI supports **35+ backends** including llama.cpp, vLLM, transformers, whisper.cpp, diffusers, MLX, MLX-VLM, and many more. Hardware acceleration is available for **NVIDIA** (CUDA 12/13), **AMD** (ROCm), **Intel** (oneAPI/SYCL), **Apple Silicon** (Metal), **Vulkan**, and **NVIDIA Jetson** (L4T). All backends can be installed on-the-fly from the [Backend Gallery](https://localai.io/backends/).
 
-### Audio & Speech Processing
-| Backend | Description | Acceleration Support |
-|---------|-------------|---------------------|
-| **whisper.cpp** | OpenAI Whisper in C/C++ | CUDA 12/13, ROCm, Intel SYCL, Vulkan, CPU |
-| **faster-whisper** | Fast Whisper with CTranslate2 | CUDA 12/13, ROCm, Intel, CPU |
-| **moonshine** | Ultra-fast transcription engine for low-end devices | CUDA 12/13, Metal, CPU |
-| **coqui** | Advanced TTS with 1100+ languages | CUDA 12/13, ROCm, Intel, CPU |
-| **kokoro** | Lightweight TTS model | CUDA 12/13, ROCm, Intel, CPU |
-| **chatterbox** | Production-grade TTS | CUDA 12/13, CPU |
-| **piper** | Fast neural TTS system | CPU |
-| **kitten-tts** | Kitten TTS models | CPU |
-| **silero-vad** | Voice Activity Detection | CPU |
-| **neutts** | Text-to-speech with voice cloning | CUDA 12/13, ROCm, CPU |
-| **vibevoice** | Real-time TTS with voice cloning | CUDA 12/13, ROCm, Intel, CPU |
-| **pocket-tts** | Lightweight CPU-based TTS | CUDA 12/13, ROCm, Intel, CPU |
-| **qwen-tts** | High-quality TTS with custom voice, voice design, and voice cloning | CUDA 12/13, ROCm, Intel, CPU |
-| **ace-step** | Music generation from text descriptions, lyrics, or audio samples | CUDA 12/13, ROCm, Intel, Metal, CPU |
+See the full [Backend & Model Compatibility Table](https://localai.io/model-compatibility/) and [GPU Acceleration guide](https://localai.io/features/gpu-acceleration/).
 
-### Image & Video Generation
-| Backend | Description | Acceleration Support |
-|---------|-------------|---------------------|
-| **stablediffusion.cpp** | Stable Diffusion in C/C++ | CUDA 12/13, Intel SYCL, Vulkan, CPU |
-| **diffusers** | HuggingFace diffusion models | CUDA 12/13, ROCm, Intel, Metal, CPU |
+## Resources
 
-### Specialized AI Tasks
-| Backend | Description | Acceleration Support |
-|---------|-------------|---------------------|
-| **rfdetr** | Real-time object detection | CUDA 12/13, Intel, CPU |
-| **rerankers** | Document reranking API | CUDA 12/13, ROCm, Intel, CPU |
-| **local-store** | Vector database | CPU |
-| **huggingface** | HuggingFace API integration | API-based |
+- [Documentation](https://localai.io/)
+- [LLM fine-tuning guide](https://localai.io/docs/advanced/fine-tuning/)
+- [Build from source](https://localai.io/basics/build/)
+- [Kubernetes installation](https://localai.io/basics/getting_started/#run-localai-in-kubernetes)
+- [Integrations & community projects](https://localai.io/docs/integrations/)
+- [Media & blog posts](https://localai.io/basics/news/#media-blogs-social)
+- [Examples](https://github.com/mudler/LocalAI-examples)
 
-### Hardware Acceleration Matrix
+## Autonomous Development Team
 
-| Acceleration Type | Supported Backends | Hardware Support |
-|-------------------|-------------------|------------------|
-| **NVIDIA CUDA 12** | All CUDA-compatible backends | Nvidia hardware |
-| **NVIDIA CUDA 13** | All CUDA-compatible backends | Nvidia hardware |
-| **AMD ROCm** | llama.cpp, whisper, vllm, transformers, diffusers, rerankers, coqui, kokoro, neutts, vibevoice, pocket-tts, qwen-tts, ace-step | AMD Graphics |
-| **Intel oneAPI** | llama.cpp, whisper, stablediffusion, vllm, transformers, diffusers, rfdetr, rerankers, coqui, kokoro, vibevoice, pocket-tts, qwen-tts, ace-step | Intel Arc, Intel iGPUs |
-| **Apple Metal** | llama.cpp, whisper, diffusers, MLX, MLX-VLM, moonshine, ace-step | Apple M1/M2/M3+ |
-| **Vulkan** | llama.cpp, whisper, stablediffusion | Cross-platform GPUs |
-| **NVIDIA Jetson (CUDA 12)** | llama.cpp, whisper, stablediffusion, diffusers, rfdetr, ace-step | ARM64 embedded AI (AGX Orin, etc.) |
-| **NVIDIA Jetson (CUDA 13)** | llama.cpp, whisper, stablediffusion, diffusers, rfdetr | ARM64 embedded AI (DGX Spark) |
-| **CPU Optimized** | All backends | AVX/AVX2/AVX512, quantization support |
+LocalAI is helped being maintained by a team of autonomous AI agents led by an AI Scrum Master.
 
-### 🔗 Community and integrations
-
-Build and deploy custom containers:
-- https://github.com/sozercan/aikit
-
-WebUIs:
-- https://github.com/Jirubizu/localai-admin
-- https://github.com/go-skynet/LocalAI-frontend
-- QA-Pilot(An interactive chat project that leverages LocalAI LLMs for rapid understanding and navigation of GitHub code repository) https://github.com/reid41/QA-Pilot
-
-Agentic Libraries:
-- https://github.com/mudler/cogito
-
-MCPs:
-- https://github.com/mudler/MCPs
-
-OS Assistant:
-
-- https://github.com/mudler/Keygeist - Keygeist is an AI-powered keyboard operator that listens for key combinations and responds with AI-generated text typed directly into your Linux box.
-
-Model galleries
-- https://github.com/go-skynet/model-gallery
-
-Voice:
-- https://github.com/richiejp/VoxInput
-
-Other:
-- Helm chart https://github.com/go-skynet/helm-charts
-- VSCode extension https://github.com/badgooooor/localai-vscode-plugin
-- Langchain: https://python.langchain.com/docs/integrations/providers/localai/
-- Terminal utility https://github.com/djcopley/ShellOracle
-- Local Smart assistant https://github.com/mudler/LocalAGI
-- Home Assistant https://github.com/sammcj/homeassistant-localai / https://github.com/drndos/hass-openai-custom-conversation / https://github.com/valentinfrlch/ha-gpt4vision
-- Discord bot https://github.com/mudler/LocalAGI/tree/main/examples/discord
-- Slack bot https://github.com/mudler/LocalAGI/tree/main/examples/slack
-- Shell-Pilot(Interact with LLM using LocalAI models via pure shell scripts on your Linux or MacOS system) https://github.com/reid41/shell-pilot
-- Telegram bot https://github.com/mudler/LocalAI/tree/master/examples/telegram-bot
-- Another Telegram Bot https://github.com/JackBekket/Hellper
-- Auto-documentation https://github.com/JackBekket/Reflexia
-- Github bot which answer on issues, with code and documentation as context https://github.com/JackBekket/GitHelper
-- Github Actions: https://github.com/marketplace/actions/start-localai
-- Examples: https://github.com/mudler/LocalAI/tree/master/examples/
-  
-
-### 🔗 Resources
-
-- [LLM finetuning guide](https://localai.io/docs/advanced/fine-tuning/)
-- [How to build locally](https://localai.io/basics/build/index.html)
-- [How to install in Kubernetes](https://localai.io/basics/getting_started/index.html#run-localai-in-kubernetes)
-- [Projects integrating LocalAI](https://localai.io/docs/integrations/)
-- [How tos section](https://io.midori-ai.xyz/howtos/) (curated by our community)
-
-## :book: 🎥 [Media, Blogs, Social](https://localai.io/basics/news/#media-blogs-social)
-
-- [Run Visual studio code with LocalAI (SUSE)](https://www.suse.com/c/running-ai-locally/)
-- 🆕 [Run LocalAI on Jetson Nano Devkit](https://mudler.pm/posts/local-ai-jetson-nano-devkit/)
-- [Run LocalAI on AWS EKS with Pulumi](https://www.pulumi.com/blog/low-code-llm-apps-with-local-ai-flowise-and-pulumi/)
-- [Run LocalAI on AWS](https://staleks.hashnode.dev/installing-localai-on-aws-ec2-instance)
-- [Create a slackbot for teams and OSS projects that answer to documentation](https://mudler.pm/posts/smart-slackbot-for-teams/)
-- [LocalAI meets k8sgpt](https://www.youtube.com/watch?v=PKrDNuJ_dfE)
-- [Question Answering on Documents locally with LangChain, LocalAI, Chroma, and GPT4All](https://mudler.pm/posts/localai-question-answering/)
-- [Tutorial to use k8sgpt with LocalAI](https://medium.com/@tyler_97636/k8sgpt-localai-unlock-kubernetes-superpowers-for-free-584790de9b65)
+- **Live Reports**: [reports.localai.io](http://reports.localai.io)
+- **Project Board**: [Agent task tracking](https://github.com/users/mudler/projects/6)
+- **Blog Post**: [Learn about the experiment](https://mudler.pm/posts/2026/02/28/a-call-to-open-source-maintainers-stop-babysitting-ai-how-i-built-a-100-local-autonomous-dev-team-to-maintain-localai-and-why-you-should-too/)
 
 ## Citation
 
@@ -388,7 +221,7 @@ If you utilize this repository, data in a downstream project, please consider ci
   howpublished = {\url{https://github.com/go-skynet/LocalAI}},
 ```
 
-## ❤️ Sponsors
+## Sponsors
 
 > Do you find LocalAI useful?
 
@@ -407,19 +240,19 @@ A huge thank you to our generous sponsors who support this project covering CI e
 
 ### Individual sponsors
 
-A special thanks to individual sponsors that contributed to the project, a full list is in [Github](https://github.com/sponsors/mudler) and [buymeacoffee](https://buymeacoffee.com/mudler), a special shout out goes to [drikster80](https://github.com/drikster80) for being generous. Thank you everyone!
+A special thanks to individual sponsors, a full list is on [GitHub](https://github.com/sponsors/mudler) and [buymeacoffee](https://buymeacoffee.com/mudler). Special shout out to [drikster80](https://github.com/drikster80) for being generous. Thank you everyone!
 
-## 🌟 Star history
+## Star history
 
 [![LocalAI Star history Chart](https://api.star-history.com/svg?repos=go-skynet/LocalAI&type=Date)](https://star-history.com/#go-skynet/LocalAI&Date)
 
-## 📖 License
+## License
 
 LocalAI is a community-driven project created by [Ettore Di Giacinto](https://github.com/mudler/).
 
 MIT - Author Ettore Di Giacinto <mudler@localai.io>
 
-## 🙇 Acknowledgements
+## Acknowledgements
 
 LocalAI couldn't have been built without the help of great software already available from the community. Thank you!
 
@@ -430,10 +263,11 @@ LocalAI couldn't have been built without the help of great software already avai
 - https://github.com/EdVince/Stable-Diffusion-NCNN
 - https://github.com/ggerganov/whisper.cpp
 - https://github.com/rhasspy/piper
+- [exo](https://github.com/exo-explore/exo) for the MLX distributed auto-parallel sharding implementation
 
-## 🤗 Contributors
+## Contributors
 
-This is a community project, a special thanks to our contributors! 🤗
+This is a community project, a special thanks to our contributors!
 <a href="https://github.com/go-skynet/LocalAI/graphs/contributors">
   <img src="https://contrib.rocks/image?repo=go-skynet/LocalAI" />
 </a>

SECURITY.md

@@ -8,10 +8,24 @@ At LocalAI, we take the security of our software seriously. We understand the im
 
 We provide support and updates for certain versions of our software. The following table outlines which versions are currently supported with security updates:
 
-| Version | Supported          |
-| ------- | ------------------ |
-| > 2.0   | :white_check_mark: |
-| < 2.0   | :x:                |
+| Version Series | Support Level | Details |
+| -------------- | ------------- | ------- |
+| 3.x | :white_check_mark: Actively supported | Full security updates and bug fixes for the latest minor versions. |
+| 2.x | :warning: Security fixes only | Critical security patches only, until **December 31, 2025**. |
+| 1.x | :x: End-of-life (EOL) | No longer supported as of **January 1, 2024**. No security fixes will be provided. |
+
+### What each support level means
+
+- **Actively supported (3.x):** Receives all security updates, bug fixes, and new features. Users should stay on the latest 3.x minor release for the best protection.
+- **Security fixes only (2.x):** Receives only critical security patches (e.g., remote code execution, authentication bypass, data exposure). No bug fixes or new features. Support ends December 31, 2025.
+- **End-of-life (1.x):** No updates of any kind. Users on 1.x are strongly encouraged to upgrade immediately, as known vulnerabilities will not be patched.
+
+### Migrating from older versions
+
+If you are running an unsupported or soon-to-be-unsupported version, we recommend upgrading as soon as possible:
+
+- **From 1.x to 3.x:** Version 1.x reached end-of-life on January 1, 2024. Review the [release notes](https://github.com/mudler/LocalAI/releases) for breaking changes across major versions, and upgrade directly to the latest 3.x release.
+- **From 2.x to 3.x:** While 2.x still receives critical security patches until December 31, 2025, we recommend planning your migration to 3.x to benefit from ongoing improvements and full support.
 
 Please ensure that you are using a supported version to receive the latest security updates.
 

aio/cpu/README.md

@@ -1,5 +0,0 @@
-## AIO CPU size
-
-Use this image with CPU-only.
-
-Please keep using only C++ backends so the base image is as small as possible (without CUDA, cuDNN, python, etc).

aio/cpu/embeddings.yaml

@@ -1,13 +0,0 @@
-embeddings: true
-name: text-embedding-ada-002
-backend: llama-cpp
-parameters:
-  model: huggingface://bartowski/granite-embedding-107m-multilingual-GGUF/granite-embedding-107m-multilingual-f16.gguf
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/cpu/rerank.yaml

@@ -1,33 +0,0 @@
-name: jina-reranker-v1-base-en
-reranking: true
-f16: true
-parameters:
-  model: jina-reranker-v1-tiny-en.f16.gguf
-backend: llama-cpp
-download_files:
-  - filename: jina-reranker-v1-tiny-en.f16.gguf
-    sha256: 5f696cf0d0f3d347c4a279eee8270e5918554cdac0ed1f632f2619e4e8341407
-    uri: huggingface://mradermacher/jina-reranker-v1-tiny-en-GGUF/jina-reranker-v1-tiny-en.f16.gguf 
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/cpu/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/cpu/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-backend: piper
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"voice-en-us-amy-low",
-      "input": "Hi, this is a test."
-    }'

aio/cpu/vad.yaml

@@ -1,8 +0,0 @@
-backend: silero-vad
-name: silero-vad
-parameters:
-  model: silero-vad.onnx
-download_files:
-- filename: silero-vad.onnx
-  uri: https://huggingface.co/onnx-community/silero-vad/resolve/main/onnx/model.onnx
-  sha256: a4a068cd6cf1ea8355b84327595838ca748ec29a25bc91fc82e6c299ccdc5808

aio/cpu/vision.yaml

@@ -1,50 +0,0 @@
-context_size: 4096
-f16: true
-backend: llama-cpp
-mmap: true
-mmproj: minicpm-v-4_5-mmproj-f16.gguf
-name: gpt-4o
-parameters:
-  model: minicpm-v-4_5-Q4_K_M.gguf
-stopwords:
-- <|im_end|>
-- <dummy32000>
-- </s>
-- <|endoftext|>
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{ .RoleName }}
-    {{ if .FunctionCall -}}
-    Function call:
-    {{ else if eq .RoleName "tool" -}}
-    Function response:
-    {{ end -}}
-    {{ if .Content -}}
-    {{.Content }}
-    {{ end -}}
-    {{ if .FunctionCall -}}
-    {{toJson .FunctionCall}}
-    {{ end -}}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |
-    <|im_start|>system
-    You are a function calling AI model. You are provided with functions to execute. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools:
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    For each function call return a json object with function name and arguments
-    <|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant
-
-download_files:
-- filename: minicpm-v-4_5-Q4_K_M.gguf
-  sha256: c1c3c33100b15b4caf7319acce4e23c0eb0ce1cbd12f70e8d24f05aa67b7512f
-  uri: huggingface://openbmb/MiniCPM-V-4_5-gguf/ggml-model-Q4_K_M.gguf
-- filename: minicpm-v-4_5-mmproj-f16.gguf
-  uri: huggingface://openbmb/MiniCPM-V-4_5-gguf/mmproj-model-f16.gguf
-  sha256: 7a7225a32e8d453aaa3d22d8c579b5bf833c253f784cdb05c99c9a76fd616df8

aio/entrypoint.sh

@@ -1,138 +0,0 @@
-#!/bin/bash
-
-echo "===> LocalAI All-in-One (AIO) container starting..."
-
-GPU_ACCELERATION=false
-GPU_VENDOR=""
-
-function check_intel() {
-    if lspci | grep -E 'VGA|3D' | grep -iq intel; then
-        echo "Intel GPU detected"
-        if [ -d /opt/intel ]; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=intel
-        else
-            echo "Intel GPU detected, but Intel GPU drivers are not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_nvidia_wsl() {
-    if lspci | grep -E 'VGA|3D' | grep -iq "Microsoft Corporation Device 008e"; then
-        # We make the assumption this WSL2 cars is NVIDIA, then check for nvidia-smi
-        # Make sure the container was run with `--gpus all` as the only required parameter
-        echo "NVIDIA GPU detected via WSL2"
-        # nvidia-smi should be installed in the container
-        if nvidia-smi; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=nvidia
-        else
-            echo "NVIDIA GPU detected via WSL2, but nvidia-smi is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_amd() {
-    if lspci | grep -E 'VGA|3D' | grep -iq amd; then
-        echo "AMD GPU detected"
-        # Check if ROCm is installed
-        if [ -d /opt/rocm ]; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=amd
-        else
-            echo "AMD GPU detected, but ROCm is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_nvidia() {
-    if lspci | grep -E 'VGA|3D' | grep -iq nvidia; then
-        echo "NVIDIA GPU detected"
-        # nvidia-smi should be installed in the container
-        if nvidia-smi; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=nvidia
-        else
-            echo "NVIDIA GPU detected, but nvidia-smi is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_metal() {
-    if system_profiler SPDisplaysDataType | grep -iq 'Metal'; then
-        echo "Apple Metal supported GPU detected"
-        GPU_ACCELERATION=true
-        GPU_VENDOR=apple
-    fi
-}
-
-function detect_gpu() {
-    case "$(uname -s)" in
-        Linux)
-            check_nvidia
-            check_amd
-            check_intel
-            check_nvidia_wsl
-            ;;
-        Darwin)
-            check_metal
-            ;;
-    esac
-}
-
-function detect_gpu_size() {
-    # Attempting to find GPU memory size for NVIDIA GPUs
-    if [ "$GPU_ACCELERATION" = true ] && [ "$GPU_VENDOR" = "nvidia" ]; then
-        echo "NVIDIA GPU detected. Attempting to find memory size..."
-        # Using head -n 1 to get the total memory of the 1st NVIDIA GPU detected.
-        # If handling multiple GPUs is required in the future, this is the place to do it
-        nvidia_sm=$(nvidia-smi --query-gpu=memory.total --format=csv,noheader,nounits | head -n 1)
-        if [ ! -z "$nvidia_sm" ]; then
-            echo "Total GPU Memory: $nvidia_sm MiB"
-            # if bigger than 8GB, use 16GB
-            #if [ "$nvidia_sm" -gt 8192 ]; then
-            #    GPU_SIZE=gpu-16g
-            #else
-            GPU_SIZE=gpu-8g
-            #fi
-        else
-            echo "Unable to determine NVIDIA GPU memory size. Falling back to CPU."
-            GPU_SIZE=gpu-8g
-        fi
-    elif [ "$GPU_ACCELERATION" = true ] && [ "$GPU_VENDOR" = "intel" ]; then
-        GPU_SIZE=intel
-    # Default to a generic GPU size until we implement GPU size detection for non NVIDIA GPUs
-    elif [ "$GPU_ACCELERATION" = true ]; then
-        echo "Non-NVIDIA GPU detected. Specific GPU memory size detection is not implemented."
-        GPU_SIZE=gpu-8g
-
-    # default to cpu if GPU_SIZE is not set
-    else
-        echo "GPU acceleration is not enabled or supported. Defaulting to CPU."
-        GPU_SIZE=cpu
-    fi
-}
-
-function check_vars() {
-    if [ -z "$MODELS" ]; then
-        echo "MODELS environment variable is not set. Please set it to a comma-separated list of model YAML files to load."
-        exit 1
-    fi
-
-    if [ -z "$PROFILE" ]; then
-        echo "PROFILE environment variable is not set. Please set it to one of the following: cpu, gpu-8g, gpu-16g, apple"
-        exit 1
-    fi
-}
-
-detect_gpu
-detect_gpu_size
-
-PROFILE="${PROFILE:-$GPU_SIZE}" # default to cpu
-export MODELS="${MODELS:-/aio/${PROFILE}/embeddings.yaml,/aio/${PROFILE}/rerank.yaml,/aio/${PROFILE}/text-to-speech.yaml,/aio/${PROFILE}/image-gen.yaml,/aio/${PROFILE}/text-to-text.yaml,/aio/${PROFILE}/speech-to-text.yaml,/aio/${PROFILE}/vad.yaml,/aio/${PROFILE}/vision.yaml}"
-
-check_vars
-
-echo "===> Starting LocalAI[$PROFILE] with the following models: $MODELS"
-
-exec /entrypoint.sh "$@"

aio/gpu-8g/embeddings.yaml

@@ -1,13 +0,0 @@
-embeddings: true
-name: text-embedding-ada-002
-backend: llama-cpp
-parameters:
-  model: huggingface://bartowski/granite-embedding-107m-multilingual-GGUF/granite-embedding-107m-multilingual-f16.gguf
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/gpu-8g/image-gen.yaml

@@ -1,25 +0,0 @@
-name: stablediffusion
-parameters:
-  model: DreamShaper_8_pruned.safetensors
-backend: diffusers
-step: 25
-f16: true
-
-diffusers:
-  pipeline_type: StableDiffusionPipeline
-  cuda: true
-  enable_parameters: "negative_prompt,num_inference_steps"
-  scheduler_type: "k_dpmpp_2m"
-
-download_files:
-- filename: DreamShaper_8_pruned.safetensors
-  uri: huggingface://Lykon/DreamShaper/DreamShaper_8_pruned.safetensors
-
-usage: |
-        curl http://localhost:8080/v1/images/generations \
-          -H "Content-Type: application/json" \
-          -d '{
-            "prompt": "<positive prompt>|<negative prompt>",
-            "step": 25,
-            "size": "512x512"
-          }'

aio/gpu-8g/rerank.yaml

@@ -1,33 +0,0 @@
-name: jina-reranker-v1-base-en
-reranking: true
-f16: true
-parameters:
-  model: jina-reranker-v1-tiny-en.f16.gguf
-backend: llama-cpp
-download_files:
-  - filename: jina-reranker-v1-tiny-en.f16.gguf
-    sha256: 5f696cf0d0f3d347c4a279eee8270e5918554cdac0ed1f632f2619e4e8341407
-    uri: huggingface://mradermacher/jina-reranker-v1-tiny-en-GGUF/jina-reranker-v1-tiny-en.f16.gguf 
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/gpu-8g/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/gpu-8g/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-backend: piper
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"tts-1",
-      "input": "Hi, this is a test."
-    }'

aio/gpu-8g/text-to-text.yaml

@@ -1,54 +0,0 @@
-context_size: 4096
-f16: true
-backend: llama-cpp
-function:
-  capture_llm_results:
-  - (?s)<Thought>(.*?)</Thought>
-  grammar:
-    properties_order: name,arguments
-  json_regex_match:
-  - (?s)<Output>(.*?)</Output>
-  replace_llm_results:
-  - key: (?s)<Thought>(.*?)</Thought>
-    value: ""
-mmap: true
-name: gpt-4
-parameters:
-  model: localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf
-stopwords:
-- <|im_end|>
-- <dummy32000>
-- </s>
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{ .RoleName }}
-    {{ if .FunctionCall -}}
-    Function call:
-    {{ else if eq .RoleName "tool" -}}
-    Function response:
-    {{ end -}}
-    {{ if .Content -}}
-    {{.Content }}
-    {{ end -}}
-    {{ if .FunctionCall -}}
-    {{toJson .FunctionCall}}
-    {{ end -}}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |
-    <|im_start|>system
-    You are an AI assistant that executes function calls, and these are the tools at your disposal:
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    <|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant
-
-download_files:
-- filename: localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf
-  sha256: 4e7b7fe1d54b881f1ef90799219dc6cc285d29db24f559c8998d1addb35713d4
-  uri: huggingface://mudler/LocalAI-functioncall-qwen2.5-7b-v0.5-Q4_K_M-GGUF/localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf

aio/gpu-8g/vad.yaml

@@ -1,8 +0,0 @@
-backend: silero-vad
-name: silero-vad
-parameters:
-  model: silero-vad.onnx
-download_files:
-- filename: silero-vad.onnx
-  uri: https://huggingface.co/onnx-community/silero-vad/resolve/main/onnx/model.onnx
-  sha256: a4a068cd6cf1ea8355b84327595838ca748ec29a25bc91fc82e6c299ccdc5808

aio/gpu-8g/vision.yaml

@@ -1,50 +0,0 @@
-context_size: 4096
-backend: llama-cpp
-f16: true
-mmap: true
-mmproj: minicpm-v-4_5-mmproj-f16.gguf
-name: gpt-4o
-parameters:
-  model: minicpm-v-4_5-Q4_K_M.gguf
-stopwords:
-- <|im_end|>
-- <dummy32000>
-- </s>
-- <|endoftext|>
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{ .RoleName }}
-    {{ if .FunctionCall -}}
-    Function call:
-    {{ else if eq .RoleName "tool" -}}
-    Function response:
-    {{ end -}}
-    {{ if .Content -}}
-    {{.Content }}
-    {{ end -}}
-    {{ if .FunctionCall -}}
-    {{toJson .FunctionCall}}
-    {{ end -}}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |
-    <|im_start|>system
-    You are a function calling AI model. You are provided with functions to execute. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools:
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    For each function call return a json object with function name and arguments
-    <|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant
-
-download_files:
-- filename: minicpm-v-4_5-Q4_K_M.gguf
-  sha256: c1c3c33100b15b4caf7319acce4e23c0eb0ce1cbd12f70e8d24f05aa67b7512f
-  uri: huggingface://openbmb/MiniCPM-V-4_5-gguf/ggml-model-Q4_K_M.gguf
-- filename: minicpm-v-4_5-mmproj-f16.gguf
-  uri: huggingface://openbmb/MiniCPM-V-4_5-gguf/mmproj-model-f16.gguf
-  sha256: 7a7225a32e8d453aaa3d22d8c579b5bf833c253f784cdb05c99c9a76fd616df8

aio/intel/embeddings.yaml

@@ -1,13 +0,0 @@
-embeddings: true
-name: text-embedding-ada-002
-backend: llama-cpp
-parameters:
-  model: huggingface://bartowski/granite-embedding-107m-multilingual-GGUF/granite-embedding-107m-multilingual-f16.gguf
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/intel/image-gen.yaml

@@ -1,20 +0,0 @@
-name: stablediffusion
-parameters:
-  model: Lykon/dreamshaper-8
-backend: diffusers
-step: 25
-f16: true
-diffusers:
-  pipeline_type: StableDiffusionPipeline
-  cuda: true
-  enable_parameters: "negative_prompt,num_inference_steps"
-  scheduler_type: "k_dpmpp_2m"
-
-usage: |
-        curl http://localhost:8080/v1/images/generations \
-          -H "Content-Type: application/json" \
-          -d '{
-            "prompt": "<positive prompt>|<negative prompt>",
-            "step": 25,
-            "size": "512x512"
-          }'

aio/intel/rerank.yaml

@@ -1,33 +0,0 @@
-name: jina-reranker-v1-base-en
-reranking: true
-f16: true
-parameters:
-  model: jina-reranker-v1-tiny-en.f16.gguf
-backend: llama-cpp
-download_files:
-  - filename: jina-reranker-v1-tiny-en.f16.gguf
-    sha256: 5f696cf0d0f3d347c4a279eee8270e5918554cdac0ed1f632f2619e4e8341407
-    uri: huggingface://mradermacher/jina-reranker-v1-tiny-en-GGUF/jina-reranker-v1-tiny-en.f16.gguf 
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/intel/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/intel/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-backend: piper
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"tts-1",
-      "input": "Hi, this is a test."
-    }'

aio/intel/text-to-text.yaml

@@ -1,54 +0,0 @@
-context_size: 4096
-f16: true
-backend: llama-cpp
-function:
-  capture_llm_results:
-  - (?s)<Thought>(.*?)</Thought>
-  grammar:
-    properties_order: name,arguments
-  json_regex_match:
-  - (?s)<Output>(.*?)</Output>
-  replace_llm_results:
-  - key: (?s)<Thought>(.*?)</Thought>
-    value: ""
-mmap: true
-name: gpt-4
-parameters:
-  model: localai-functioncall-qwen2.5-7b-v0.5-q4_k_m.gguf
-stopwords:
-- <|im_end|>
-- <dummy32000>
-- </s>
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{ .RoleName }}
-    {{ if .FunctionCall -}}
-    Function call:
-    {{ else if eq .RoleName "tool" -}}
-    Function response:
-    {{ end -}}
-    {{ if .Content -}}
-    {{.Content }}
-    {{ end -}}
-    {{ if .FunctionCall -}}
-    {{toJson .FunctionCall}}
-    {{ end -}}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |
-    <|im_start|>system
-    You are an AI assistant that executes function calls, and these are the tools at your disposal:
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    <|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant
-
-download_files:
-- filename: localai-functioncall-phi-4-v0.3-q4_k_m.gguf
-  sha256: 23fee048ded2a6e2e1a7b6bbefa6cbf83068f194caa9552aecbaa00fec8a16d5
-  uri: huggingface://mudler/LocalAI-functioncall-phi-4-v0.3-Q4_K_M-GGUF/localai-functioncall-phi-4-v0.3-q4_k_m.gguf

backend/Dockerfile.golang

@@ -20,7 +20,7 @@ RUN apt-get update && \
         build-essential \
         git ccache \
         ca-certificates \
-        make cmake wget \
+        make cmake wget libopenblas-dev \
         curl unzip \
         libssl-dev && \
     apt-get clean && \
@@ -180,6 +180,11 @@ RUN <<EOT bash
     fi
 EOT
 
+RUN if [ "${BACKEND}" = "opus" ]; then \
+    apt-get update && apt-get install -y --no-install-recommends libopus-dev pkg-config && \
+    apt-get clean && rm -rf /var/lib/apt/lists/*; \
+fi
+
 COPY . /LocalAI
 
 RUN git config --global --add safe.directory /LocalAI

backend/Dockerfile.llama-cpp

@@ -58,7 +58,9 @@ ARG CUDA_DOCKER_ARCH
 ENV CUDA_DOCKER_ARCH=${CUDA_DOCKER_ARCH}
 ARG CMAKE_ARGS
 ENV CMAKE_ARGS=${CMAKE_ARGS}
-ARG BACKEND=rerankers
+ARG BACKEND=llama-cpp
+ARG LLAMA_BACKEND_DIR=${BACKEND}
+ENV LLAMA_BACKEND_DIR=${LLAMA_BACKEND_DIR}
 ARG BUILD_TYPE
 ENV BUILD_TYPE=${BUILD_TYPE}
 ARG CUDA_MAJOR_VERSION
@@ -255,32 +257,27 @@ if [[ -n "${CUDA_DOCKER_ARCH:-}" ]]; then
   CUDA_ARCH_ESC="${CUDA_DOCKER_ARCH//;/\\;}"
   export CMAKE_ARGS="${CMAKE_ARGS:-} -DCMAKE_CUDA_ARCHITECTURES=${CUDA_ARCH_ESC}"
   echo "CMAKE_ARGS(env) = ${CMAKE_ARGS}"
-  rm -rf /LocalAI/backend/cpp/llama-cpp-*-build
+  rm -rf /LocalAI/backend/cpp/${LLAMA_BACKEND_DIR}-*-build
 fi
 
+cd /LocalAI/backend/cpp/${LLAMA_BACKEND_DIR}
+
 if [ "${TARGETARCH}" = "arm64" ] || [ "${BUILD_TYPE}" = "hipblas" ]; then
-  cd /LocalAI/backend/cpp/llama-cpp
-  make llama-cpp-fallback
-  make llama-cpp-grpc
-  make llama-cpp-rpc-server
+  make ARCH=aarch64 build-variants
 else
-  cd /LocalAI/backend/cpp/llama-cpp
-  make llama-cpp-avx
-  make llama-cpp-avx2
-  make llama-cpp-avx512
-  make llama-cpp-fallback
-  make llama-cpp-grpc
-  make llama-cpp-rpc-server
+  make build-variants
 fi
 EOT
 
 
 # Copy libraries using a script to handle architecture differences
-RUN make -BC /LocalAI/backend/cpp/llama-cpp package
+RUN make -BC /LocalAI/backend/cpp/${LLAMA_BACKEND_DIR} package
 
 
 FROM scratch
 
+ARG BACKEND=llama-cpp
+ARG LLAMA_BACKEND_DIR=${BACKEND}
 
 # Copy all available binaries (the build process only creates the appropriate ones for the target architecture)
-COPY --from=builder /LocalAI/backend/cpp/llama-cpp/package/. ./
+COPY --from=builder /LocalAI/backend/cpp/${LLAMA_BACKEND_DIR}/package/. ./

backend/Dockerfile.python

@@ -202,6 +202,11 @@ RUN mkdir -p /${BACKEND}/lib && \
     TARGET_LIB_DIR="/${BACKEND}/lib" BUILD_TYPE="${BUILD_TYPE}" CUDA_MAJOR_VERSION="${CUDA_MAJOR_VERSION}" \
     bash /package-gpu-libs.sh "/${BACKEND}/lib"
 
+# Run backend-specific packaging if a package.sh exists
+RUN if [ -f "/${BACKEND}/package.sh" ]; then \
+        cd /${BACKEND} && bash package.sh; \
+    fi
+
 FROM scratch
 ARG BACKEND=rerankers
 COPY --from=builder /${BACKEND}/ /

backend/README.md

@@ -53,7 +53,6 @@ The backend system provides language-specific Dockerfiles that handle the build
 #### Go Backends (`go/`)
 - **whisper**: OpenAI Whisper speech recognition in Go with GGML cpp backend (whisper.cpp)
 - **stablediffusion-ggml**: Stable Diffusion in Go with GGML Cpp backend
-- **huggingface**: Hugging Face model integration
 - **piper**: Text-to-speech synthesis Golang with C bindings using rhaspy/piper
 - **local-store**: Vector storage backend
 

backend/backend.proto

@@ -9,6 +9,7 @@ package backend;
 
 service Backend {
   rpc Health(HealthMessage) returns (Reply) {}
+  rpc Free(HealthMessage) returns (Result) {}
   rpc Predict(PredictOptions) returns (Reply) {}
   rpc LoadModel(ModelOptions) returns (Result) {}
   rpc PredictStream(PredictOptions) returns (stream Reply) {}
@@ -34,7 +35,23 @@ service Backend {
 
   rpc VAD(VADRequest) returns (VADResponse) {}
 
+  rpc AudioEncode(AudioEncodeRequest) returns (AudioEncodeResult) {}
+  rpc AudioDecode(AudioDecodeRequest) returns (AudioDecodeResult) {}
+
   rpc ModelMetadata(ModelOptions) returns (ModelMetadataResponse) {}
+
+  // Fine-tuning RPCs
+  rpc StartFineTune(FineTuneRequest) returns (FineTuneJobResult) {}
+  rpc FineTuneProgress(FineTuneProgressRequest) returns (stream FineTuneProgressUpdate) {}
+  rpc StopFineTune(FineTuneStopRequest) returns (Result) {}
+  rpc ListCheckpoints(ListCheckpointsRequest) returns (ListCheckpointsResponse) {}
+  rpc ExportModel(ExportModelRequest) returns (Result) {}
+
+  // Quantization RPCs
+  rpc StartQuantization(QuantizationRequest) returns (QuantizationJobResult) {}
+  rpc QuantizationProgress(QuantizationProgressRequest) returns (stream QuantizationProgressUpdate) {}
+  rpc StopQuantization(QuantizationStopRequest) returns (Result) {}
+
 }
 
 // Define the empty request
@@ -161,6 +178,24 @@ message PredictOptions {
   string ToolChoice = 49;  // JSON string or object specifying tool choice behavior
   int32 Logprobs = 50;  // Number of top logprobs to return (maps to OpenAI logprobs parameter)
   int32 TopLogprobs = 51;  // Number of top logprobs to return per token (maps to OpenAI top_logprobs parameter)
+  map<string, string> Metadata = 52;  // Generic per-request metadata (e.g., enable_thinking)
+  float MinP = 53;  // Minimum probability sampling threshold (0.0 = disabled)
+}
+
+// ToolCallDelta represents an incremental tool call update from the C++ parser.
+// Used for both streaming (partial diffs) and non-streaming (final tool calls).
+message ToolCallDelta {
+  int32 index = 1;           // tool call index (0-based)
+  string id = 2;             // tool call ID (e.g., "call_abc123")
+  string name = 3;           // function name (set on first appearance)
+  string arguments = 4;      // arguments chunk (incremental in streaming, full in non-streaming)
+}
+
+// ChatDelta represents incremental content/reasoning/tool_call updates parsed by the C++ backend.
+message ChatDelta {
+  string content = 1;                       // content text delta
+  string reasoning_content = 2;             // reasoning/thinking text delta
+  repeated ToolCallDelta tool_calls = 3;    // tool call deltas
 }
 
 // The response message containing the result
@@ -172,6 +207,7 @@ message Reply {
   double timing_token_generation = 5;
   bytes audio = 6;
   bytes logprobs = 7;  // JSON-encoded logprobs data matching OpenAI format
+  repeated ChatDelta chat_deltas = 8;       // Parsed chat deltas from C++ autoparser (streaming + non-streaming)
 }
 
 message GrammarTrigger {
@@ -423,7 +459,222 @@ message DetectResponse {
   repeated Detection Detections = 1;
 }
 
+message ToolFormatMarkers {
+  string format_type = 1;           // "json_native", "tag_with_json", "tag_with_tagged"
+
+  // Tool section markers
+  string section_start = 2;         // e.g., "<tool_call>", "[TOOL_CALLS]"
+  string section_end = 3;           // e.g., "</tool_call>"
+  string per_call_start = 4;        // e.g., "<|tool_call_begin|>"
+  string per_call_end = 5;          // e.g., "<|tool_call_end|>"
+
+  // Function name markers (TAG_WITH_JSON / TAG_WITH_TAGGED)
+  string func_name_prefix = 6;     // e.g., "<function="
+  string func_name_suffix = 7;     // e.g., ">"
+  string func_close = 8;           // e.g., "</function>"
+
+  // Argument markers (TAG_WITH_TAGGED)
+  string arg_name_prefix = 9;      // e.g., "<param="
+  string arg_name_suffix = 10;     // e.g., ">"
+  string arg_value_prefix = 11;
+  string arg_value_suffix = 12;    // e.g., "</param>"
+  string arg_separator = 13;       // e.g., "\n"
+
+  // JSON format fields (JSON_NATIVE)
+  string name_field = 14;          // e.g., "name"
+  string args_field = 15;          // e.g., "arguments"
+  string id_field = 16;            // e.g., "id"
+  bool fun_name_is_key = 17;
+  bool tools_array_wrapped = 18;
+  reserved 19;
+
+  // Reasoning markers
+  string reasoning_start = 20;     // e.g., "<think>"
+  string reasoning_end = 21;       // e.g., "</think>"
+
+  // Content markers
+  string content_start = 22;
+  string content_end = 23;
+
+  // Args wrapper markers
+  string args_start = 24;          // e.g., "<args>"
+  string args_end = 25;            // e.g., "</args>"
+
+  // JSON parameter ordering
+  string function_field = 26;      // e.g., "function" (wrapper key in JSON)
+  repeated string parameter_order = 27;
+
+  // Generated ID field (alternative field name for generated IDs)
+  string gen_id_field = 28;        // e.g., "call_id"
+
+  // Call ID markers (position and delimiters for tool call IDs)
+  string call_id_position = 29;    // "none", "pre_func_name", "between_func_and_args", "post_args"
+  string call_id_prefix = 30;      // e.g., "[CALL_ID]"
+  string call_id_suffix = 31;      // e.g., ""
+}
+
+message AudioEncodeRequest {
+  bytes pcm_data = 1;
+  int32 sample_rate = 2;
+  int32 channels = 3;
+  map<string, string> options = 4;
+}
+
+message AudioEncodeResult {
+  repeated bytes frames = 1;
+  int32 sample_rate = 2;
+  int32 samples_per_frame = 3;
+}
+
+message AudioDecodeRequest {
+  repeated bytes frames = 1;
+  map<string, string> options = 2;
+}
+
+message AudioDecodeResult {
+  bytes pcm_data = 1;
+  int32 sample_rate = 2;
+  int32 samples_per_frame = 3;
+}
+
 message ModelMetadataResponse {
   bool supports_thinking = 1;
   string rendered_template = 2;  // The rendered chat template with enable_thinking=true (empty if not applicable)
+  ToolFormatMarkers tool_format = 3;  // Auto-detected tool format markers from differential template analysis
 }
+
+// Fine-tuning messages
+
+message FineTuneRequest {
+  // Model identification
+  string model = 1;                       // HF model name or local path
+  string training_type = 2;              // "lora", "loha", "lokr", "full" — what parameters to train
+  string training_method = 3;            // "sft", "dpo", "grpo", "rloo", "reward", "kto", "orpo", "network_training"
+
+  // Adapter config (universal across LoRA/LoHa/LoKr for LLM + diffusion)
+  int32 adapter_rank = 10;               // LoRA rank (r), default 16
+  int32 adapter_alpha = 11;              // scaling factor, default 16
+  float adapter_dropout = 12;            // default 0.0
+  repeated string target_modules = 13;   // layer names to adapt
+
+  // Universal training hyperparameters
+  float learning_rate = 20;              // default 2e-4
+  int32 num_epochs = 21;                 // default 3
+  int32 batch_size = 22;                 // default 2
+  int32 gradient_accumulation_steps = 23; // default 4
+  int32 warmup_steps = 24;              // default 5
+  int32 max_steps = 25;                 // 0 = use epochs
+  int32 save_steps = 26;               // 0 = only save final
+  float weight_decay = 27;             // default 0.01
+  bool gradient_checkpointing = 28;
+  string optimizer = 29;               // adamw_8bit, adamw, sgd, adafactor, prodigy
+  int32 seed = 30;                     // default 3407
+  string mixed_precision = 31;        // fp16, bf16, fp8, no
+
+  // Dataset
+  string dataset_source = 40;          // HF dataset ID, local file/dir path
+  string dataset_split = 41;           // train, test, etc.
+
+  // Output
+  string output_dir = 50;
+  string job_id = 51;                  // client-assigned or auto-generated
+
+  // Resume training from a checkpoint
+  string resume_from_checkpoint = 55;  // path to checkpoint dir to resume from
+
+  // Backend-specific AND method-specific extensibility
+  map<string, string> extra_options = 60;
+}
+
+message FineTuneJobResult {
+  string job_id = 1;
+  bool success = 2;
+  string message = 3;
+}
+
+message FineTuneProgressRequest {
+  string job_id = 1;
+}
+
+message FineTuneProgressUpdate {
+  string job_id = 1;
+  int32 current_step = 2;
+  int32 total_steps = 3;
+  float current_epoch = 4;
+  float total_epochs = 5;
+  float loss = 6;
+  float learning_rate = 7;
+  float grad_norm = 8;
+  float eval_loss = 9;
+  float eta_seconds = 10;
+  float progress_percent = 11;
+  string status = 12;                  // queued, caching, loading_model, loading_dataset, training, saving, completed, failed, stopped
+  string message = 13;
+  string checkpoint_path = 14;        // set when a checkpoint is saved
+  string sample_path = 15;           // set when a sample is generated (video/image backends)
+  map<string, float> extra_metrics = 16; // method-specific metrics
+}
+
+message FineTuneStopRequest {
+  string job_id = 1;
+  bool save_checkpoint = 2;
+}
+
+message ListCheckpointsRequest {
+  string output_dir = 1;
+}
+
+message ListCheckpointsResponse {
+  repeated CheckpointInfo checkpoints = 1;
+}
+
+message CheckpointInfo {
+  string path = 1;
+  int32 step = 2;
+  float epoch = 3;
+  float loss = 4;
+  string created_at = 5;
+}
+
+message ExportModelRequest {
+  string checkpoint_path = 1;
+  string output_path = 2;
+  string export_format = 3;           // lora, loha, lokr, merged_16bit, merged_4bit, gguf, diffusers
+  string quantization_method = 4;     // for GGUF: q4_k_m, q5_k_m, q8_0, f16, etc.
+  string model = 5;                   // base model name (for merge operations)
+  map<string, string> extra_options = 6;
+}
+
+// Quantization messages
+
+message QuantizationRequest {
+  string model = 1;                      // HF model name or local path
+  string quantization_type = 2;          // q4_k_m, q5_k_m, q8_0, f16, etc.
+  string output_dir = 3;                 // where to write output files
+  string job_id = 4;                     // client-assigned job ID
+  map<string, string> extra_options = 5; // hf_token, custom flags, etc.
+}
+
+message QuantizationJobResult {
+  string job_id = 1;
+  bool success = 2;
+  string message = 3;
+}
+
+message QuantizationProgressRequest {
+  string job_id = 1;
+}
+
+message QuantizationProgressUpdate {
+  string job_id = 1;
+  float progress_percent = 2;
+  string status = 3;                     // queued, downloading, converting, quantizing, completed, failed, stopped
+  string message = 4;
+  string output_file = 5;               // set when completed — path to the output GGUF file
+  map<string, float> extra_metrics = 6;  // e.g. file_size_mb, compression_ratio
+}
+
+message QuantizationStopRequest {
+  string job_id = 1;
+}
+

backend/cpp/llama-cpp-tq/Makefile

@@ -0,0 +1,6 @@
+LLAMA_VERSION?=master
+LLAMA_REPO?=https://github.com/TheTom/llama-cpp-turboquant
+BACKEND_NAME?=llama-cpp-tq
+SHARED_DIR?=$(CURDIR)/../llama-cpp
+
+include ../llama-cpp/Makefile

backend/cpp/llama-cpp/CMakeLists.txt

@@ -59,6 +59,11 @@ add_library(hw_grpc_proto
 
 add_executable(${TARGET} grpc-server.cpp json.hpp httplib.h)
 
+# Enable autoparser support if the header exists (not present in all llama.cpp forks)
+if(EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/chat-auto-parser.h")
+    target_compile_definitions(${TARGET} PRIVATE HAS_AUTOPARSER)
+endif()
+
 target_include_directories(${TARGET} PRIVATE ../llava)
 target_include_directories(${TARGET} PRIVATE ${CMAKE_SOURCE_DIR})
 

backend/cpp/llama-cpp/Makefile

@@ -1,6 +1,10 @@
 
-LLAMA_VERSION?=8872ad2125336d209a9911a82101f80095a9831d
+LLAMA_VERSION?=a1cfb645307edc61a89e41557f290f441043d3c2
 LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
+BACKEND_NAME?=llama-cpp
+SHARED_DIR?=$(CURDIR)
+GRPC_SERVER_DIR?=tools/grpc-server
+SERVER_SOURCE_DIR?=tools/server
 
 CMAKE_ARGS?=
 BUILD_TYPE?=
@@ -67,6 +71,17 @@ ifeq ($(BUILD_TYPE),sycl_f32)
 		-DCMAKE_CXX_FLAGS="-fsycl"
 endif
 
+# Variants to build for each architecture (can be overridden by forks)
+X86_64_VARIANTS ?= llama-cpp-avx llama-cpp-avx2 llama-cpp-avx512 llama-cpp-fallback llama-cpp-grpc llama-cpp-rpc-server
+ARM64_VARIANTS ?= llama-cpp-fallback llama-cpp-grpc llama-cpp-rpc-server
+
+build-variants:
+ifeq ($(ARCH),aarch64)
+	@for v in $(ARM64_VARIANTS); do $(MAKE) $$v || exit 1; done
+else
+	@for v in $(X86_64_VARIANTS); do $(MAKE) $$v || exit 1; done
+endif
+
 INSTALLED_PACKAGES=$(CURDIR)/../grpc/installed_packages
 INSTALLED_LIB_CMAKE=$(INSTALLED_PACKAGES)/lib/cmake
 ADDED_CMAKE_ARGS=-Dabsl_DIR=${INSTALLED_LIB_CMAKE}/absl \
@@ -90,42 +105,42 @@ else
 endif
 
 llama-cpp-avx2: llama.cpp
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx2-build
-	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx2-build purge
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME) $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx2-build
+	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx2-build purge
 	$(info ${GREEN}I llama-cpp build info:avx2${RESET})
-	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=on -DGGML_AVX512=off -DGGML_FMA=on -DGGML_F16C=on" $(MAKE) VARIANT="llama-cpp-avx2-build" build-llama-cpp-grpc-server
-	cp -rfv $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx2-build/grpc-server llama-cpp-avx2
+	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=on -DGGML_AVX512=off -DGGML_FMA=on -DGGML_F16C=on" $(MAKE) VARIANT="$(BACKEND_NAME)-avx2-build" build-llama-cpp-grpc-server
+	cp -rfv $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx2-build/grpc-server llama-cpp-avx2
 
 llama-cpp-avx512: llama.cpp
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx512-build
-	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx512-build purge
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME) $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx512-build
+	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx512-build purge
 	$(info ${GREEN}I llama-cpp build info:avx512${RESET})
-	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=on -DGGML_FMA=on -DGGML_F16C=on" $(MAKE) VARIANT="llama-cpp-avx512-build" build-llama-cpp-grpc-server
-	cp -rfv $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx512-build/grpc-server llama-cpp-avx512
+	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=on -DGGML_FMA=on -DGGML_F16C=on" $(MAKE) VARIANT="$(BACKEND_NAME)-avx512-build" build-llama-cpp-grpc-server
+	cp -rfv $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx512-build/grpc-server llama-cpp-avx512
 
 llama-cpp-avx: llama.cpp
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx-build
-	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx-build purge
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME) $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx-build
+	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx-build purge
 	$(info ${GREEN}I llama-cpp build info:avx${RESET})
-	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) VARIANT="llama-cpp-avx-build" build-llama-cpp-grpc-server
-	cp -rfv $(CURRENT_MAKEFILE_DIR)/../llama-cpp-avx-build/grpc-server llama-cpp-avx
+	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) VARIANT="$(BACKEND_NAME)-avx-build" build-llama-cpp-grpc-server
+	cp -rfv $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-avx-build/grpc-server llama-cpp-avx
 
 llama-cpp-fallback: llama.cpp
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp $(CURRENT_MAKEFILE_DIR)/../llama-cpp-fallback-build
-	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../llama-cpp-fallback-build purge
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME) $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-fallback-build
+	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-fallback-build purge
 	$(info ${GREEN}I llama-cpp build info:fallback${RESET})
-	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) VARIANT="llama-cpp-fallback-build" build-llama-cpp-grpc-server
-	cp -rfv $(CURRENT_MAKEFILE_DIR)/../llama-cpp-fallback-build/grpc-server llama-cpp-fallback
+	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) VARIANT="$(BACKEND_NAME)-fallback-build" build-llama-cpp-grpc-server
+	cp -rfv $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-fallback-build/grpc-server llama-cpp-fallback
 
 llama-cpp-grpc: llama.cpp
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp $(CURRENT_MAKEFILE_DIR)/../llama-cpp-grpc-build
-	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../llama-cpp-grpc-build purge
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME) $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-grpc-build
+	$(MAKE) -C $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-grpc-build purge
 	$(info ${GREEN}I llama-cpp build info:grpc${RESET})
-	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_RPC=ON -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" TARGET="--target grpc-server --target rpc-server" $(MAKE) VARIANT="llama-cpp-grpc-build" build-llama-cpp-grpc-server
-	cp -rfv $(CURRENT_MAKEFILE_DIR)/../llama-cpp-grpc-build/grpc-server llama-cpp-grpc
+	CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_RPC=ON -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" TARGET="--target grpc-server --target rpc-server" $(MAKE) VARIANT="$(BACKEND_NAME)-grpc-build" build-llama-cpp-grpc-server
+	cp -rfv $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-grpc-build/grpc-server llama-cpp-grpc
 
 llama-cpp-rpc-server: llama-cpp-grpc
-	cp -rf $(CURRENT_MAKEFILE_DIR)/../llama-cpp-grpc-build/llama.cpp/build/bin/rpc-server llama-cpp-rpc-server
+	cp -rf $(CURRENT_MAKEFILE_DIR)/../$(BACKEND_NAME)-grpc-build/llama.cpp/build/bin/rpc-server llama-cpp-rpc-server
 
 llama.cpp:
 	mkdir -p llama.cpp
@@ -133,30 +148,30 @@ llama.cpp:
 	git init && \
 	git remote add origin $(LLAMA_REPO)  && \
 	git fetch origin && \
-	git checkout -b build $(LLAMA_VERSION) && \
+	(git checkout -b build $(LLAMA_VERSION) || git checkout -b build origin/$(LLAMA_VERSION)) && \
 	git submodule update --init --recursive --depth 1 --single-branch
 
-llama.cpp/tools/grpc-server: llama.cpp
-	mkdir -p llama.cpp/tools/grpc-server
-	bash prepare.sh
+llama.cpp/$(GRPC_SERVER_DIR): llama.cpp
+	mkdir -p llama.cpp/$(GRPC_SERVER_DIR)
+	SHARED_DIR=$(SHARED_DIR) SERVER_SOURCE_DIR=$(SERVER_SOURCE_DIR) GRPC_SERVER_DIR=$(GRPC_SERVER_DIR) bash $(SHARED_DIR)/prepare.sh
 
 rebuild:
-	bash prepare.sh
+	SHARED_DIR=$(SHARED_DIR) SERVER_SOURCE_DIR=$(SERVER_SOURCE_DIR) GRPC_SERVER_DIR=$(GRPC_SERVER_DIR) bash $(SHARED_DIR)/prepare.sh
 	rm -rf grpc-server
 	$(MAKE) grpc-server
 
 package:
-	bash package.sh
+	bash $(SHARED_DIR)/package.sh
 
 purge:
 	rm -rf llama.cpp/build
-	rm -rf llama.cpp/tools/grpc-server
+	rm -rf llama.cpp/$(GRPC_SERVER_DIR)
 	rm -rf grpc-server
 
 clean: purge
 	rm -rf llama.cpp
 
-grpc-server: llama.cpp llama.cpp/tools/grpc-server
+grpc-server: llama.cpp llama.cpp/$(GRPC_SERVER_DIR)
 	@echo "Building grpc-server with $(BUILD_TYPE) build type and $(CMAKE_ARGS)"
 ifneq (,$(findstring sycl,$(BUILD_TYPE)))
 	+bash -c "source $(ONEAPI_VARS); \

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

@@ -17,12 +17,17 @@
 #include "backend.pb.h"
 #include "backend.grpc.pb.h"
 #include "common.h"
+#ifdef HAS_AUTOPARSER
+#include "chat-auto-parser.h"
+#endif
 #include <getopt.h>
 #include <grpcpp/ext/proto_server_reflection_plugin.h>
 #include <grpcpp/grpcpp.h>
 #include <grpcpp/health_check_service_interface.h>
+#include <grpcpp/security/server_credentials.h>
 #include <regex>
 #include <atomic>
+#include <cstdlib>
 #include <mutex>
 #include <signal.h>
 #include <thread>
@@ -36,6 +41,47 @@ using grpc::Server;
 using grpc::ServerBuilder;
 using grpc::ServerContext;
 using grpc::Status;
+
+// gRPC bearer token auth via AuthMetadataProcessor for distributed mode.
+// Reads LOCALAI_GRPC_AUTH_TOKEN from the environment. When set, rejects
+// requests without a matching "authorization: Bearer <token>" metadata header.
+class TokenAuthMetadataProcessor : public grpc::AuthMetadataProcessor {
+public:
+    explicit TokenAuthMetadataProcessor(const std::string& token) : token_(token) {}
+
+    bool IsBlocking() const override { return false; }
+
+    grpc::Status Process(const InputMetadata& auth_metadata,
+                         grpc::AuthContext* /*context*/,
+                         OutputMetadata* /*consumed_auth_metadata*/,
+                         OutputMetadata* /*response_metadata*/) override {
+        auto it = auth_metadata.find("authorization");
+        if (it != auth_metadata.end()) {
+            std::string expected = "Bearer " + token_;
+            std::string got(it->second.data(), it->second.size());
+            // Constant-time comparison
+            if (expected.size() == got.size() && ct_memcmp(expected.data(), got.data(), expected.size()) == 0) {
+                return grpc::Status::OK;
+            }
+        }
+        return grpc::Status(grpc::StatusCode::UNAUTHENTICATED, "invalid token");
+    }
+
+private:
+    std::string token_;
+
+    // Minimal constant-time comparison (avoids OpenSSL dependency)
+    static int ct_memcmp(const void* a, const void* b, size_t n) {
+        const unsigned char* pa = static_cast<const unsigned char*>(a);
+        const unsigned char* pb = static_cast<const unsigned char*>(b);
+        unsigned char result = 0;
+        for (size_t i = 0; i < n; i++) {
+            result |= pa[i] ^ pb[i];
+        }
+        return result;
+    }
+};
+
 // END LocalAI
 
 
@@ -135,6 +181,7 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, const
     data["mirostat_eta"] = predict->mirostateta();
     data["n_keep"] = predict->nkeep();
     data["seed"] = predict->seed();
+    data["min_p"] = predict->minp();
 
 
     std::string grammar_str = predict->grammar();
@@ -362,7 +409,7 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
       params.mmproj.path = request->mmproj();
     }
     //  params.model_alias ??
-    params.model_alias =  request->modelfile();
+    params.model_alias.insert(request->modelfile());
     if (!request->cachetypekey().empty()) {
         params.cache_type_k = kv_cache_type_from_str(request->cachetypekey());
     }
@@ -417,6 +464,12 @@ static void params_parse(server_context& /*ctx_server*/, const backend::ModelOpt
     // n_ctx_checkpoints: max context checkpoints per slot (default: 8)
     params.n_ctx_checkpoints = 8;
 
+    // llama memory fit fails if we don't provide a buffer for tensor overrides
+    const size_t ntbo = llama_max_tensor_buft_overrides();
+    while (params.tensor_buft_overrides.size() < ntbo) {
+        params.tensor_buft_overrides.push_back({nullptr, nullptr});
+    }
+
      // decode options. Options are in form optname:optvale, or if booleans only optname.
     for (int i = 0; i < request->options_size(); i++) {
         std::string opt = request->options(i);
@@ -860,6 +913,56 @@ public:
         return logprobs_json;
     }
 
+    // Helper: populate chat_deltas on a Reply from oaicompat_msg_diffs (streaming chunks)
+    static void populate_chat_deltas_from_diffs(backend::Reply & reply,
+                                                const std::vector<common_chat_msg_diff> & diffs) {
+        for (const auto & diff : diffs) {
+            auto* delta = reply.add_chat_deltas();
+            if (!diff.content_delta.empty()) {
+                delta->set_content(diff.content_delta);
+            }
+            if (!diff.reasoning_content_delta.empty()) {
+                delta->set_reasoning_content(diff.reasoning_content_delta);
+            }
+            if (diff.tool_call_index != std::string::npos) {
+                auto* tc = delta->add_tool_calls();
+                tc->set_index(static_cast<int32_t>(diff.tool_call_index));
+                if (!diff.tool_call_delta.id.empty()) {
+                    tc->set_id(diff.tool_call_delta.id);
+                }
+                if (!diff.tool_call_delta.name.empty()) {
+                    tc->set_name(diff.tool_call_delta.name);
+                }
+                if (!diff.tool_call_delta.arguments.empty()) {
+                    tc->set_arguments(diff.tool_call_delta.arguments);
+                }
+            }
+        }
+    }
+
+    // Helper: populate chat_deltas on a Reply from final oaicompat_msg (non-streaming)
+    static void populate_chat_deltas_from_final(backend::Reply & reply,
+                                                const common_chat_msg & msg) {
+        // Content delta
+        if (!msg.content.empty() || !msg.reasoning_content.empty() || !msg.tool_calls.empty()) {
+            auto* delta = reply.add_chat_deltas();
+            if (!msg.content.empty()) {
+                delta->set_content(msg.content);
+            }
+            if (!msg.reasoning_content.empty()) {
+                delta->set_reasoning_content(msg.reasoning_content);
+            }
+            // Tool calls as individual deltas within the same ChatDelta
+            for (size_t i = 0; i < msg.tool_calls.size(); i++) {
+                auto* tc = delta->add_tool_calls();
+                tc->set_index(static_cast<int32_t>(i));
+                tc->set_id(msg.tool_calls[i].id);
+                tc->set_name(msg.tool_calls[i].name);
+                tc->set_arguments(msg.tool_calls[i].arguments);
+            }
+        }
+    }
+
     grpc::Status PredictStream(grpc::ServerContext* context, const backend::PredictOptions* request, grpc::ServerWriter<backend::Reply>* writer) override {
         if (params_base.model.path.empty()) {
             return grpc::Status(grpc::StatusCode::FAILED_PRECONDITION, "Model not loaded");
@@ -1255,6 +1358,52 @@ public:
                     body_json["add_generation_prompt"] = data["add_generation_prompt"];
                 }
 
+                // Pass sampling parameters to body_json so oaicompat_chat_params_parse respects them
+                // and doesn't overwrite them with defaults in the returned parsed_data
+                if (data.contains("n_predict")) {
+                    body_json["max_tokens"] = data["n_predict"];
+                }
+                if (data.contains("ignore_eos")) {
+                    body_json["ignore_eos"] = data["ignore_eos"];
+                }
+                if (data.contains("stop")) {
+                    body_json["stop"] = data["stop"];
+                }
+                if (data.contains("temperature")) {
+                    body_json["temperature"] = data["temperature"];
+                }
+                if (data.contains("top_p")) {
+                    body_json["top_p"] = data["top_p"];
+                }
+                if (data.contains("frequency_penalty")) {
+                    body_json["frequency_penalty"] = data["frequency_penalty"];
+                }
+                if (data.contains("presence_penalty")) {
+                    body_json["presence_penalty"] = data["presence_penalty"];
+                }
+                if (data.contains("seed")) {
+                    body_json["seed"] = data["seed"];
+                }
+                if (data.contains("logit_bias")) {
+                    body_json["logit_bias"] = data["logit_bias"];
+                }
+                if (data.contains("top_k")) {
+                    body_json["top_k"] = data["top_k"];
+                }
+                if (data.contains("min_p")) {
+                    body_json["min_p"] = data["min_p"];
+                }
+
+                // Pass enable_thinking via chat_template_kwargs (where oaicompat_chat_params_parse reads it)
+                const auto& metadata = request->metadata();
+                auto et_it = metadata.find("enable_thinking");
+                if (et_it != metadata.end()) {
+                    if (!body_json.contains("chat_template_kwargs")) {
+                        body_json["chat_template_kwargs"] = json::object();
+                    }
+                    body_json["chat_template_kwargs"]["enable_thinking"] = (et_it->second == "true");
+                }
+
                 // Debug: Print full body_json before template processing (includes messages, tools, tool_choice, etc.)
                 SRV_DBG("[CONVERSATION DEBUG] PredictStream: Full body_json before oaicompat_chat_params_parse:\n%s\n", body_json.dump(2).c_str());
 
@@ -1435,127 +1584,76 @@ public:
             return grpc::Status(grpc::StatusCode::INTERNAL, error_json.value("message", "Error occurred"));
         }
 
+        // Lambda to build a Reply from JSON + attach chat deltas from a result
+        auto build_reply_from_json = [](const json & res_json, server_task_result * raw_result) -> backend::Reply {
+            backend::Reply reply;
+            std::string completion_text = res_json.value("content", "");
+            reply.set_message(completion_text);
+            reply.set_tokens(res_json.value("tokens_predicted", 0));
+            reply.set_prompt_tokens(res_json.value("tokens_evaluated", 0));
+
+            if (res_json.contains("timings")) {
+                reply.set_timing_prompt_processing(res_json.at("timings").value("prompt_ms", 0.0));
+                reply.set_timing_token_generation(res_json.at("timings").value("predicted_ms", 0.0));
+            }
+
+            json logprobs_json = extract_logprobs_from_json(res_json);
+            if (!logprobs_json.empty() && !logprobs_json.is_null()) {
+                reply.set_logprobs(logprobs_json.dump());
+            }
+
+            return reply;
+        };
+
+        auto attach_chat_deltas = [](backend::Reply & reply, server_task_result * raw_result) {
+            // Try streaming partial result first
+            auto* partial = dynamic_cast<server_task_result_cmpl_partial*>(raw_result);
+            if (partial && !partial->oaicompat_msg_diffs.empty()) {
+                populate_chat_deltas_from_diffs(reply, partial->oaicompat_msg_diffs);
+                return;
+            }
+            // Try final result
+            auto* final_res = dynamic_cast<server_task_result_cmpl_final*>(raw_result);
+            if (final_res && final_res->is_updated) {
+                populate_chat_deltas_from_diffs(reply, final_res->oaicompat_msg_diffs);
+            }
+        };
+
         // Process first result
         json first_res_json = first_result->to_json();
         if (first_res_json.is_array()) {
             for (const auto & res : first_res_json) {
-                std::string completion_text = res.value("content", "");
-
-                backend::Reply reply;
-                reply.set_message(completion_text);
-                int32_t tokens_predicted = res.value("tokens_predicted", 0);
-                reply.set_tokens(tokens_predicted);
-                int32_t tokens_evaluated = res.value("tokens_evaluated", 0);
-                reply.set_prompt_tokens(tokens_evaluated);
-
-                if (res.contains("timings")) {
-                    double timing_prompt_processing = res.at("timings").value("prompt_ms", 0.0);
-                    reply.set_timing_prompt_processing(timing_prompt_processing);
-                    double timing_token_generation = res.at("timings").value("predicted_ms", 0.0);
-                    reply.set_timing_token_generation(timing_token_generation);
-                }
-
-                // Extract and set logprobs if present
-                json logprobs_json = extract_logprobs_from_json(res);
-                if (!logprobs_json.empty() && !logprobs_json.is_null()) {
-                    std::string logprobs_str = logprobs_json.dump();
-                    reply.set_logprobs(logprobs_str);
-                }
-
+                auto reply = build_reply_from_json(res, first_result.get());
+                attach_chat_deltas(reply, first_result.get());
                 writer->Write(reply);
             }
         } else {
-            std::string completion_text = first_res_json.value("content", "");
-
-            backend::Reply reply;
-            reply.set_message(completion_text);
-            int32_t tokens_predicted = first_res_json.value("tokens_predicted", 0);
-            reply.set_tokens(tokens_predicted);
-            int32_t tokens_evaluated = first_res_json.value("tokens_evaluated", 0);
-            reply.set_prompt_tokens(tokens_evaluated);
-
-            if (first_res_json.contains("timings")) {
-                double timing_prompt_processing = first_res_json.at("timings").value("prompt_ms", 0.0);
-                reply.set_timing_prompt_processing(timing_prompt_processing);
-                double timing_token_generation = first_res_json.at("timings").value("predicted_ms", 0.0);
-                reply.set_timing_token_generation(timing_token_generation);
-            }
-
-            // Extract and set logprobs if present
-            json logprobs_json = extract_logprobs_from_json(first_res_json);
-            if (!logprobs_json.empty() && !logprobs_json.is_null()) {
-                std::string logprobs_str = logprobs_json.dump();
-                reply.set_logprobs(logprobs_str);
-            }
-
+            auto reply = build_reply_from_json(first_res_json, first_result.get());
+            attach_chat_deltas(reply, first_result.get());
             writer->Write(reply);
         }
 
         // Process subsequent results
         while (rd.has_next()) {
-            // Check if context is cancelled before processing result
             if (context->IsCancelled()) {
                 break;
             }
 
             auto result = rd.next([&context]() { return context->IsCancelled(); });
             if (result == nullptr) {
-                // connection is closed
                 break;
             }
 
             json res_json = result->to_json();
             if (res_json.is_array()) {
                 for (const auto & res : res_json) {
-                    std::string completion_text = res.value("content", "");
-
-                    backend::Reply reply;
-                    reply.set_message(completion_text);
-                    int32_t tokens_predicted = res.value("tokens_predicted", 0);
-                    reply.set_tokens(tokens_predicted);
-                    int32_t tokens_evaluated = res.value("tokens_evaluated", 0);
-                    reply.set_prompt_tokens(tokens_evaluated);
-
-                    if (res.contains("timings")) {
-                        double timing_prompt_processing = res.at("timings").value("prompt_ms", 0.0);
-                        reply.set_timing_prompt_processing(timing_prompt_processing);
-                        double timing_token_generation = res.at("timings").value("predicted_ms", 0.0);
-                        reply.set_timing_token_generation(timing_token_generation);
-                    }
-
-                    // Extract and set logprobs if present
-                    json logprobs_json = extract_logprobs_from_json(res);
-                    if (!logprobs_json.empty() && !logprobs_json.is_null()) {
-                        std::string logprobs_str = logprobs_json.dump();
-                        reply.set_logprobs(logprobs_str);
-                    }
-
+                    auto reply = build_reply_from_json(res, result.get());
+                    attach_chat_deltas(reply, result.get());
                     writer->Write(reply);
                 }
             } else {
-                std::string completion_text = res_json.value("content", "");
-
-                backend::Reply reply;
-                reply.set_message(completion_text);
-                int32_t tokens_predicted = res_json.value("tokens_predicted", 0);
-                reply.set_tokens(tokens_predicted);
-                int32_t tokens_evaluated = res_json.value("tokens_evaluated", 0);
-                reply.set_prompt_tokens(tokens_evaluated);
-
-                if (res_json.contains("timings")) {
-                    double timing_prompt_processing = res_json.at("timings").value("prompt_ms", 0.0);
-                    reply.set_timing_prompt_processing(timing_prompt_processing);
-                    double timing_token_generation = res_json.at("timings").value("predicted_ms", 0.0);
-                    reply.set_timing_token_generation(timing_token_generation);
-                }
-
-                // Extract and set logprobs if present
-                json logprobs_json = extract_logprobs_from_json(res_json);
-                if (!logprobs_json.empty() && !logprobs_json.is_null()) {
-                    std::string logprobs_str = logprobs_json.dump();
-                    reply.set_logprobs(logprobs_str);
-                }
-
+                auto reply = build_reply_from_json(res_json, result.get());
+                attach_chat_deltas(reply, result.get());
                 writer->Write(reply);
             }
         }
@@ -1986,6 +2084,52 @@ public:
                     body_json["add_generation_prompt"] = data["add_generation_prompt"];
                 }
 
+                // Pass sampling parameters to body_json so oaicompat_chat_params_parse respects them
+                // and doesn't overwrite them with defaults in the returned parsed_data
+                if (data.contains("n_predict")) {
+                    body_json["max_tokens"] = data["n_predict"];
+                }
+                if (data.contains("ignore_eos")) {
+                    body_json["ignore_eos"] = data["ignore_eos"];
+                }
+                if (data.contains("stop")) {
+                    body_json["stop"] = data["stop"];
+                }
+                if (data.contains("temperature")) {
+                    body_json["temperature"] = data["temperature"];
+                }
+                if (data.contains("top_p")) {
+                    body_json["top_p"] = data["top_p"];
+                }
+                if (data.contains("frequency_penalty")) {
+                    body_json["frequency_penalty"] = data["frequency_penalty"];
+                }
+                if (data.contains("presence_penalty")) {
+                    body_json["presence_penalty"] = data["presence_penalty"];
+                }
+                if (data.contains("seed")) {
+                    body_json["seed"] = data["seed"];
+                }
+                if (data.contains("logit_bias")) {
+                    body_json["logit_bias"] = data["logit_bias"];
+                }
+                if (data.contains("top_k")) {
+                    body_json["top_k"] = data["top_k"];
+                }
+                if (data.contains("min_p")) {
+                    body_json["min_p"] = data["min_p"];
+                }
+
+                // Pass enable_thinking via chat_template_kwargs (where oaicompat_chat_params_parse reads it)
+                const auto& predict_metadata = request->metadata();
+                auto predict_et_it = predict_metadata.find("enable_thinking");
+                if (predict_et_it != predict_metadata.end()) {
+                    if (!body_json.contains("chat_template_kwargs")) {
+                        body_json["chat_template_kwargs"] = json::object();
+                    }
+                    body_json["chat_template_kwargs"]["enable_thinking"] = (predict_et_it->second == "true");
+                }
+
                 // Debug: Print full body_json before template processing (includes messages, tools, tool_choice, etc.)
                 SRV_DBG("[CONVERSATION DEBUG] Predict: Full body_json before oaicompat_chat_params_parse:\n%s\n", body_json.dump(2).c_str());
 
@@ -2172,7 +2316,8 @@ public:
             std::cout << "[DEBUG] Received " << all_results.results.size() << " results" << std::endl;
             if (all_results.results.size() == 1) {
                 // single result
-                GGML_ASSERT(dynamic_cast<server_task_result_cmpl_final*>(all_results.results[0].get()) != nullptr);
+                auto* final_res = dynamic_cast<server_task_result_cmpl_final*>(all_results.results[0].get());
+                GGML_ASSERT(final_res != nullptr);
                 json result_json = all_results.results[0]->to_json();
                 reply->set_message(result_json.value("content", ""));
 
@@ -2195,6 +2340,11 @@ public:
                     reply->set_logprobs(logprobs_str);
                 }
 
+                // Populate chat deltas from the autoparser's final parsed message
+                if (final_res->is_updated) {
+                    populate_chat_deltas_from_final(*reply, final_res->oaicompat_msg);
+                }
+
             } else {
                 // multiple results (multitask)
                 json arr = json::array();
@@ -2517,6 +2667,114 @@ public:
         
         response->set_rendered_template(rendered_template);
 
+#ifdef HAS_AUTOPARSER
+        // Run differential template analysis to detect tool format markers
+        if (params_base.use_jinja) {
+            try {
+                // Get template source and reconstruct a common_chat_template for analysis
+                std::string tmpl_src = common_chat_templates_source(ctx_server.impl->chat_params.tmpls.get());
+                if (!tmpl_src.empty()) {
+                    const auto * vocab = llama_model_get_vocab(ctx_server.impl->model);
+                    std::string token_bos, token_eos;
+                    if (vocab) {
+                        auto bos_id = llama_vocab_bos(vocab);
+                        auto eos_id = llama_vocab_eos(vocab);
+                        if (bos_id != LLAMA_TOKEN_NULL) {
+                            token_bos = common_token_to_piece(vocab, bos_id, true);
+                        }
+                        if (eos_id != LLAMA_TOKEN_NULL) {
+                            token_eos = common_token_to_piece(vocab, eos_id, true);
+                        }
+                    }
+                    common_chat_template tmpl(tmpl_src, token_bos, token_eos);
+                    struct autoparser::autoparser ap;
+                    ap.analyze_template(tmpl);
+
+                    if (ap.analysis_complete && ap.tools.format.mode != autoparser::tool_format::NONE) {
+                        auto * tf = response->mutable_tool_format();
+
+                        // Format type
+                        switch (ap.tools.format.mode) {
+                            case autoparser::tool_format::JSON_NATIVE:
+                                tf->set_format_type("json_native");
+                                break;
+                            case autoparser::tool_format::TAG_WITH_JSON:
+                                tf->set_format_type("tag_with_json");
+                                break;
+                            case autoparser::tool_format::TAG_WITH_TAGGED:
+                                tf->set_format_type("tag_with_tagged");
+                                break;
+                            default:
+                                break;
+                        }
+
+                        // Tool section markers
+                        tf->set_section_start(ap.tools.format.section_start);
+                        tf->set_section_end(ap.tools.format.section_end);
+                        tf->set_per_call_start(ap.tools.format.per_call_start);
+                        tf->set_per_call_end(ap.tools.format.per_call_end);
+
+                        // Function markers
+                        tf->set_func_name_prefix(ap.tools.function.name_prefix);
+                        tf->set_func_name_suffix(ap.tools.function.name_suffix);
+                        tf->set_func_close(ap.tools.function.close);
+
+                        // Argument markers
+                        tf->set_arg_name_prefix(ap.tools.arguments.name_prefix);
+                        tf->set_arg_name_suffix(ap.tools.arguments.name_suffix);
+                        tf->set_arg_value_prefix(ap.tools.arguments.value_prefix);
+                        tf->set_arg_value_suffix(ap.tools.arguments.value_suffix);
+                        tf->set_arg_separator(ap.tools.arguments.separator);
+                        tf->set_args_start(ap.tools.arguments.start);
+                        tf->set_args_end(ap.tools.arguments.end);
+
+                        // JSON format fields
+                        tf->set_name_field(ap.tools.format.name_field);
+                        tf->set_args_field(ap.tools.format.args_field);
+                        tf->set_id_field(ap.tools.format.id_field);
+                        tf->set_fun_name_is_key(ap.tools.format.fun_name_is_key);
+                        tf->set_tools_array_wrapped(ap.tools.format.tools_array_wrapped);
+                        tf->set_function_field(ap.tools.format.function_field);
+
+                        tf->set_gen_id_field(ap.tools.format.gen_id_field);
+
+                        for (const auto & p : ap.tools.format.parameter_order) {
+                            tf->add_parameter_order(p);
+                        }
+
+                        // Call ID markers
+                        switch (ap.tools.call_id.pos) {
+                            case autoparser::call_id_position::NONE:
+                                tf->set_call_id_position("none");
+                                break;
+                            case autoparser::call_id_position::PRE_FUNC_NAME:
+                                tf->set_call_id_position("pre_func_name");
+                                break;
+                            case autoparser::call_id_position::BETWEEN_FUNC_AND_ARGS:
+                                tf->set_call_id_position("between_func_and_args");
+                                break;
+                            case autoparser::call_id_position::POST_ARGS:
+                                tf->set_call_id_position("post_args");
+                                break;
+                        }
+                        tf->set_call_id_prefix(ap.tools.call_id.prefix);
+                        tf->set_call_id_suffix(ap.tools.call_id.suffix);
+
+                        // Reasoning markers
+                        tf->set_reasoning_start(ap.reasoning.start);
+                        tf->set_reasoning_end(ap.reasoning.end);
+
+                        // Content markers
+                        tf->set_content_start(ap.content.start);
+                        tf->set_content_end(ap.content.end);
+                    }
+                }
+            } catch (const std::exception & e) {
+                SRV_WRN("ModelMetadata: failed to run autoparser analysis: %s\n", e.what());
+            }
+        }
+#endif
+
         return grpc::Status::OK;
     }
 };
@@ -2549,11 +2807,24 @@ int main(int argc, char** argv) {
     BackendServiceImpl service(ctx_server);
 
     ServerBuilder builder;
-    builder.AddListeningPort(server_address, grpc::InsecureServerCredentials());
+    // Add bearer token auth via AuthMetadataProcessor if LOCALAI_GRPC_AUTH_TOKEN is set
+    const char* auth_token = std::getenv("LOCALAI_GRPC_AUTH_TOKEN");
+    std::shared_ptr<grpc::ServerCredentials> creds;
+    if (auth_token != nullptr && auth_token[0] != '\0') {
+        creds = grpc::InsecureServerCredentials();
+        creds->SetAuthMetadataProcessor(
+            std::make_shared<TokenAuthMetadataProcessor>(auth_token));
+        std::cout << "gRPC auth enabled via LOCALAI_GRPC_AUTH_TOKEN" << std::endl;
+    } else {
+        creds = grpc::InsecureServerCredentials();
+    }
+
+    builder.AddListeningPort(server_address, creds);
     builder.RegisterService(&service);
     builder.SetMaxMessageSize(50 * 1024 * 1024); // 50MB
     builder.SetMaxSendMessageSize(50 * 1024 * 1024); // 50MB
     builder.SetMaxReceiveMessageSize(50 * 1024 * 1024); // 50MB
+
     std::unique_ptr<Server> server(builder.BuildAndStart());
    // run the HTTP server in a thread - see comment below
     std::thread t([&]()

backend/cpp/llama-cpp/package.sh

@@ -5,14 +5,21 @@
 
 set -e
 
-CURDIR=$(dirname "$(realpath $0)")
-REPO_ROOT="${CURDIR}/../../.."
+# Use working directory (not script location) so forks that share this script work correctly
+CURDIR=$(pwd)
+SCRIPT_DIR=$(dirname "$(realpath $0)")
+REPO_ROOT="${SCRIPT_DIR}/../../.."
 
 # Create lib directory
 mkdir -p $CURDIR/package/lib
 
 cp -avrf $CURDIR/llama-cpp-* $CURDIR/package/
-cp -rfv $CURDIR/run.sh $CURDIR/package/
+# Copy run.sh — prefer local copy, fall back to shared dir (script location)
+if [ -f "$CURDIR/run.sh" ]; then
+    cp -rfv $CURDIR/run.sh $CURDIR/package/
+else
+    cp -rfv $SCRIPT_DIR/run.sh $CURDIR/package/
+fi
 
 # Detect architecture and copy appropriate libraries
 if [ -f "/lib64/ld-linux-x86-64.so.2" ]; then
@@ -24,6 +31,9 @@ if [ -f "/lib64/ld-linux-x86-64.so.2" ]; then
     cp -arfLv /lib/x86_64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
     cp -arfLv /lib/x86_64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
     cp -arfLv /lib/x86_64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libdl.so.2 $CURDIR/package/lib/libdl.so.2
+    cp -arfLv /lib/x86_64-linux-gnu/librt.so.1 $CURDIR/package/lib/librt.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libpthread.so.0 $CURDIR/package/lib/libpthread.so.0
 elif [ -f "/lib/ld-linux-aarch64.so.1" ]; then
     # ARM64 architecture
     echo "Detected ARM64 architecture, copying ARM64 libraries..."
@@ -33,6 +43,9 @@ elif [ -f "/lib/ld-linux-aarch64.so.1" ]; then
     cp -arfLv /lib/aarch64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
     cp -arfLv /lib/aarch64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
     cp -arfLv /lib/aarch64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libdl.so.2 $CURDIR/package/lib/libdl.so.2
+    cp -arfLv /lib/aarch64-linux-gnu/librt.so.1 $CURDIR/package/lib/librt.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libpthread.so.0 $CURDIR/package/lib/libpthread.so.0
 else
     echo "Error: Could not detect architecture"
     exit 1

backend/cpp/llama-cpp/prepare.sh

@@ -1,31 +1,43 @@
 #!/bin/bash
 
-## Patches
+SHARED_DIR="${SHARED_DIR:-.}"
+SERVER_SOURCE_DIR="${SERVER_SOURCE_DIR:-tools/server}"
+GRPC_SERVER_DIR="${GRPC_SERVER_DIR:-tools/grpc-server}"
 
 ## Apply patches from the `patches` directory
 if [ -d "patches" ]; then
     for patch in $(ls patches); do
         echo "Applying patch $patch"
         patch -d llama.cpp/ -p1 < patches/$patch
-    done 
+    done
 fi
 
 set -e
 
-for file in $(ls llama.cpp/tools/server/); do
-    cp -rfv llama.cpp/tools/server/$file llama.cpp/tools/grpc-server/
+# Copy server source files into grpc-server build directory
+for file in $(ls llama.cpp/${SERVER_SOURCE_DIR}/); do
+    cp -rfv llama.cpp/${SERVER_SOURCE_DIR}/$file llama.cpp/${GRPC_SERVER_DIR}/
 done
 
-cp -r CMakeLists.txt llama.cpp/tools/grpc-server/
-cp -r grpc-server.cpp llama.cpp/tools/grpc-server/
-cp -rfv llama.cpp/vendor/nlohmann/json.hpp llama.cpp/tools/grpc-server/
-cp -rfv llama.cpp/vendor/cpp-httplib/httplib.h llama.cpp/tools/grpc-server/
+# Copy build files — prefer local overrides, fall back to SHARED_DIR
+for f in CMakeLists.txt grpc-server.cpp; do
+    if [ -f "$f" ]; then
+        cp -r "$f" llama.cpp/${GRPC_SERVER_DIR}/
+    else
+        cp -r "$SHARED_DIR/$f" llama.cpp/${GRPC_SERVER_DIR}/
+    fi
+done
+
+cp -rfv llama.cpp/vendor/nlohmann/json.hpp llama.cpp/${GRPC_SERVER_DIR}/
+cp -rfv llama.cpp/vendor/cpp-httplib/httplib.h llama.cpp/${GRPC_SERVER_DIR}/
+
+# Add grpc-server subdirectory to the parent CMakeLists.txt
+PARENT_CMAKELISTS="llama.cpp/$(dirname ${GRPC_SERVER_DIR})/CMakeLists.txt"
 
 set +e
-if grep -q "grpc-server" llama.cpp/tools/CMakeLists.txt; then
+if grep -q "grpc-server" "$PARENT_CMAKELISTS"; then
     echo "grpc-server already added"
 else
-    echo "add_subdirectory(grpc-server)" >> llama.cpp/tools/CMakeLists.txt
+    echo "add_subdirectory(grpc-server)" >> "$PARENT_CMAKELISTS"
 fi
 set -e
-

backend/go/acestep-cpp/CMakeLists.txt

@@ -0,0 +1,54 @@
+cmake_minimum_required(VERSION 3.14)
+project(goacestepcpp LANGUAGES C CXX)
+set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
+
+set(ACESTEP_DIR ${CMAKE_CURRENT_SOURCE_DIR}/sources/acestep.cpp)
+
+# Override upstream's CMAKE_CUDA_ARCHITECTURES before add_subdirectory.
+# Upstream sets 120a/121a for CUDA >= 12.8, but those archs require a newer
+# toolkit than 12.8.x ships. Pre-defining this variable makes the upstream
+# "if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES)" guard skip its broken defaults.
+if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES)
+    set(CMAKE_CUDA_ARCHITECTURES "75-virtual;80-virtual;86-real;89-real")
+endif()
+
+# EXCLUDE_FROM_ALL: only build targets we explicitly depend on (acestep-core, ggml),
+# skip upstream standalone executables (ace-understand, dit-vae, etc.)
+add_subdirectory(${ACESTEP_DIR} acestep EXCLUDE_FROM_ALL)
+
+add_library(goacestepcpp MODULE cpp/goacestepcpp.cpp)
+target_link_libraries(goacestepcpp PRIVATE acestep-core ggml ggml-base ggml-cpu)
+
+# Include dirs matching link_ggml_backends macro, but with absolute paths
+target_include_directories(goacestepcpp PRIVATE ${ACESTEP_DIR}/src ${ACESTEP_DIR})
+target_include_directories(goacestepcpp SYSTEM PRIVATE ${ACESTEP_DIR}/ggml/include)
+
+# Link GPU backends if available (mirrors link_ggml_backends macro)
+foreach(backend blas cuda metal vulkan)
+    if(TARGET ggml-${backend})
+        target_link_libraries(goacestepcpp PRIVATE ggml-${backend})
+        string(TOUPPER ${backend} BACKEND_UPPER)
+        target_compile_definitions(goacestepcpp PRIVATE ACESTEP_HAVE_${BACKEND_UPPER})
+        if(backend STREQUAL "cuda")
+            find_package(CUDAToolkit QUIET)
+            if(CUDAToolkit_FOUND)
+                target_link_libraries(goacestepcpp PRIVATE CUDA::cudart)
+            endif()
+        endif()
+    endif()
+endforeach()
+
+if(MSVC)
+    target_compile_options(goacestepcpp PRIVATE /W4 /wd4100 /wd4505)
+else()
+    target_compile_options(goacestepcpp PRIVATE -Wall -Wextra -Wshadow -Wconversion
+                          -Wno-unused-parameter -Wno-unused-function -Wno-sign-conversion)
+endif()
+
+if(CMAKE_CXX_COMPILER_ID MATCHES "GNU" AND CMAKE_CXX_COMPILER_VERSION VERSION_LESS 9.0)
+    target_link_libraries(goacestepcpp PRIVATE stdc++fs)
+endif()
+
+set_property(TARGET goacestepcpp PROPERTY CXX_STANDARD 17)
+set_target_properties(goacestepcpp PROPERTIES LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR})

backend/go/acestep-cpp/Makefile

@@ -0,0 +1,127 @@
+CMAKE_ARGS?=
+BUILD_TYPE?=
+NATIVE?=false
+
+GOCMD?=go
+GO_TAGS?=
+JOBS?=$(shell nproc --ignore=1)
+
+# acestep.cpp version
+ACESTEP_REPO?=https://github.com/ace-step/acestep.cpp
+ACESTEP_CPP_VERSION?=6f35c874ee11e86d511b860019b84976f5b52d3a
+SO_TARGET?=libgoacestepcpp.so
+
+CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF
+
+ifeq ($(NATIVE),false)
+	CMAKE_ARGS+=-DGGML_NATIVE=OFF
+endif
+
+ifeq ($(BUILD_TYPE),cublas)
+	CMAKE_ARGS+=-DGGML_CUDA=ON
+else ifeq ($(BUILD_TYPE),openblas)
+	CMAKE_ARGS+=-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS
+else ifeq ($(BUILD_TYPE),clblas)
+	CMAKE_ARGS+=-DGGML_CLBLAST=ON -DCLBlast_DIR=/some/path
+else ifeq ($(BUILD_TYPE),hipblas)
+	CMAKE_ARGS+=-DGGML_HIPBLAS=ON
+else ifeq ($(BUILD_TYPE),vulkan)
+	CMAKE_ARGS+=-DGGML_VULKAN=ON
+else ifeq ($(OS),Darwin)
+	ifneq ($(BUILD_TYPE),metal)
+		CMAKE_ARGS+=-DGGML_METAL=OFF
+	else
+		CMAKE_ARGS+=-DGGML_METAL=ON
+		CMAKE_ARGS+=-DGGML_METAL_EMBED_LIBRARY=ON
+	endif
+endif
+
+ifeq ($(BUILD_TYPE),sycl_f16)
+	CMAKE_ARGS+=-DGGML_SYCL=ON \
+		-DCMAKE_C_COMPILER=icx \
+		-DCMAKE_CXX_COMPILER=icpx \
+		-DGGML_SYCL_F16=ON
+endif
+
+ifeq ($(BUILD_TYPE),sycl_f32)
+	CMAKE_ARGS+=-DGGML_SYCL=ON \
+		-DCMAKE_C_COMPILER=icx \
+		-DCMAKE_CXX_COMPILER=icpx
+endif
+
+sources/acestep.cpp:
+	mkdir -p sources/acestep.cpp
+	cd sources/acestep.cpp && \
+	git init && \
+	git remote add origin $(ACESTEP_REPO) && \
+	git fetch origin && \
+	git checkout $(ACESTEP_CPP_VERSION) && \
+	git submodule update --init --recursive --depth 1 --single-branch
+
+# Detect OS
+UNAME_S := $(shell uname -s)
+
+# Only build CPU variants on Linux
+ifeq ($(UNAME_S),Linux)
+	VARIANT_TARGETS = libgoacestepcpp-avx.so libgoacestepcpp-avx2.so libgoacestepcpp-avx512.so libgoacestepcpp-fallback.so
+else
+	# On non-Linux (e.g., Darwin), build only fallback variant
+	VARIANT_TARGETS = libgoacestepcpp-fallback.so
+endif
+
+acestep-cpp: main.go goacestepcpp.go $(VARIANT_TARGETS)
+	CGO_ENABLED=0 $(GOCMD) build -tags "$(GO_TAGS)" -o acestep-cpp ./
+
+package: acestep-cpp
+	bash package.sh
+
+build: package
+
+clean: purge
+	rm -rf libgoacestepcpp*.so package sources/acestep.cpp acestep-cpp
+
+purge:
+	rm -rf build*
+
+# Variants must build sequentially: each uses its own build-<name> directory,
+# but parallel builds can still race on shared resources (jobserver, disk I/O).
+.NOTPARALLEL:
+
+# Build all variants (Linux only)
+ifeq ($(UNAME_S),Linux)
+libgoacestepcpp-avx.so: sources/acestep.cpp
+	$(info ${GREEN}I acestep-cpp build info:avx${RESET})
+	SO_TARGET=libgoacestepcpp-avx.so CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) libgoacestepcpp-custom
+	rm -rf build-libgoacestepcpp-avx.so
+
+libgoacestepcpp-avx2.so: sources/acestep.cpp
+	$(info ${GREEN}I acestep-cpp build info:avx2${RESET})
+	SO_TARGET=libgoacestepcpp-avx2.so CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=on -DGGML_AVX512=off -DGGML_FMA=on -DGGML_F16C=on -DGGML_BMI2=on" $(MAKE) libgoacestepcpp-custom
+	rm -rf build-libgoacestepcpp-avx2.so
+
+libgoacestepcpp-avx512.so: sources/acestep.cpp
+	$(info ${GREEN}I acestep-cpp build info:avx512${RESET})
+	SO_TARGET=libgoacestepcpp-avx512.so CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=on -DGGML_AVX2=on -DGGML_AVX512=on -DGGML_FMA=on -DGGML_F16C=on -DGGML_BMI2=on" $(MAKE) libgoacestepcpp-custom
+	rm -rf build-libgoacestepcpp-avx512.so
+endif
+
+# Build fallback variant (all platforms)
+libgoacestepcpp-fallback.so: sources/acestep.cpp
+	$(info ${GREEN}I acestep-cpp build info:fallback${RESET})
+	SO_TARGET=libgoacestepcpp-fallback.so CMAKE_ARGS="$(CMAKE_ARGS) -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_BMI2=off" $(MAKE) libgoacestepcpp-custom
+	rm -rf build-libgoacestepcpp-fallback.so
+
+libgoacestepcpp-custom: CMakeLists.txt cpp/goacestepcpp.cpp cpp/goacestepcpp.h
+	mkdir -p build-$(SO_TARGET) && \
+	cd build-$(SO_TARGET) && \
+	cmake .. $(CMAKE_ARGS) && \
+	cmake --build . --config Release -j$(JOBS) --target goacestepcpp && \
+	cd .. && \
+	mv build-$(SO_TARGET)/libgoacestepcpp.so ./$(SO_TARGET)
+
+test: acestep-cpp
+	@echo "Running acestep-cpp tests..."
+	bash test.sh
+	@echo "acestep-cpp tests completed."
+
+all: acestep-cpp package

backend/go/acestep-cpp/acestepcpp_test.go

@@ -0,0 +1,204 @@
+package main
+
+import (
+	"context"
+	"os"
+	"os/exec"
+	"path/filepath"
+	"testing"
+	"time"
+
+	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
+	"google.golang.org/grpc"
+	"google.golang.org/grpc/credentials/insecure"
+)
+
+const (
+	testAddr    = "localhost:50051"
+	startupWait = 5 * time.Second
+)
+
+func skipIfNoModel(t *testing.T) string {
+	t.Helper()
+	modelDir := os.Getenv("ACESTEP_MODEL_DIR")
+	if modelDir == "" {
+		t.Skip("ACESTEP_MODEL_DIR not set, skipping test (set to directory with GGUF models)")
+	}
+	if _, err := os.Stat(filepath.Join(modelDir, "acestep-5Hz-lm-0.6B-Q8_0.gguf")); os.IsNotExist(err) {
+		t.Skipf("LM model file not found in %s, skipping", modelDir)
+	}
+	if _, err := os.Stat(filepath.Join(modelDir, "Qwen3-Embedding-0.6B-Q8_0.gguf")); os.IsNotExist(err) {
+		t.Skipf("Text encoder model file not found in %s, skipping", modelDir)
+	}
+	if _, err := os.Stat(filepath.Join(modelDir, "acestep-v15-turbo-Q8_0.gguf")); os.IsNotExist(err) {
+		t.Skipf("DiT model file not found in %s, skipping", modelDir)
+	}
+	if _, err := os.Stat(filepath.Join(modelDir, "vae-BF16.gguf")); os.IsNotExist(err) {
+		t.Skipf("VAE model file not found in %s, skipping", modelDir)
+	}
+	return modelDir
+}
+
+func startServer(t *testing.T) *exec.Cmd {
+	t.Helper()
+	binary := os.Getenv("ACESTEP_BINARY")
+	if binary == "" {
+		binary = "./acestep-cpp"
+	}
+	if _, err := os.Stat(binary); os.IsNotExist(err) {
+		t.Skipf("Backend binary not found at %s, skipping", binary)
+	}
+	cmd := exec.Command(binary, "--addr", testAddr)
+	cmd.Stdout = os.Stderr
+	cmd.Stderr = os.Stderr
+	if err := cmd.Start(); err != nil {
+		t.Fatalf("Failed to start server: %v", err)
+	}
+	time.Sleep(startupWait)
+	return cmd
+}
+
+func stopServer(cmd *exec.Cmd) {
+	if cmd != nil && cmd.Process != nil {
+		cmd.Process.Kill()
+		cmd.Wait()
+	}
+}
+
+func dialGRPC(t *testing.T) *grpc.ClientConn {
+	t.Helper()
+	conn, err := grpc.Dial(testAddr,
+		grpc.WithTransportCredentials(insecure.NewCredentials()),
+		grpc.WithDefaultCallOptions(
+			grpc.MaxCallRecvMsgSize(50*1024*1024),
+			grpc.MaxCallSendMsgSize(50*1024*1024),
+		),
+	)
+	if err != nil {
+		t.Fatalf("Failed to dial gRPC: %v", err)
+	}
+	return conn
+}
+
+func TestServerHealth(t *testing.T) {
+	cmd := startServer(t)
+	defer stopServer(cmd)
+
+	conn := dialGRPC(t)
+	defer conn.Close()
+
+	client := pb.NewBackendClient(conn)
+	resp, err := client.Health(context.Background(), &pb.HealthMessage{})
+	if err != nil {
+		t.Fatalf("Health check failed: %v", err)
+	}
+	if string(resp.Message) != "OK" {
+		t.Fatalf("Expected OK, got %s", string(resp.Message))
+	}
+}
+
+func TestLoadModel(t *testing.T) {
+	modelDir := skipIfNoModel(t)
+	cmd := startServer(t)
+	defer stopServer(cmd)
+
+	conn := dialGRPC(t)
+	defer conn.Close()
+
+	client := pb.NewBackendClient(conn)
+
+	// Get base directory from main model file for relative paths
+	mainModelPath := filepath.Join(modelDir, "acestep-5Hz-lm-0.6B-Q8_0.gguf")
+
+	resp, err := client.LoadModel(context.Background(), &pb.ModelOptions{
+		ModelFile: mainModelPath,
+		ModelPath: modelDir,
+		Options: []string{
+			"text_encoder_model:Qwen3-Embedding-0.6B-Q8_0.gguf",
+			"dit_model:acestep-v15-turbo-Q8_0.gguf",
+			"vae_model:vae-BF16.gguf",
+		},
+	})
+	if err != nil {
+		t.Fatalf("LoadModel failed: %v", err)
+	}
+	if !resp.Success {
+		t.Fatalf("LoadModel returned failure: %s", resp.Message)
+	}
+}
+
+func TestSoundGeneration(t *testing.T) {
+	modelDir := skipIfNoModel(t)
+
+	tmpDir, err := os.MkdirTemp("", "acestep-test")
+	if err != nil {
+		t.Fatal(err)
+	}
+	t.Cleanup(func() { os.RemoveAll(tmpDir) })
+
+	outputFile := filepath.Join(tmpDir, "output.wav")
+
+	cmd := startServer(t)
+	defer stopServer(cmd)
+
+	conn := dialGRPC(t)
+	defer conn.Close()
+
+	client := pb.NewBackendClient(conn)
+
+	// Get base directory from main model file for relative paths
+	mainModelPath := filepath.Join(modelDir, "acestep-5Hz-lm-0.6B-Q8_0.gguf")
+
+	// Load models
+	loadResp, err := client.LoadModel(context.Background(), &pb.ModelOptions{
+		ModelFile: mainModelPath,
+		ModelPath: modelDir,
+		Options: []string{
+			"text_encoder_model:Qwen3-Embedding-0.6B-Q8_0.gguf",
+			"dit_model:acestep-v15-turbo-Q8_0.gguf",
+			"vae_model:vae-BF16.gguf",
+		},
+	})
+	if err != nil {
+		t.Fatalf("LoadModel failed: %v", err)
+	}
+	if !loadResp.Success {
+		t.Fatalf("LoadModel returned failure: %s", loadResp.Message)
+	}
+
+	// Generate music
+	duration := float32(10.0)
+	temperature := float32(0.85)
+	bpm := int32(120)
+	caption := "A cheerful electronic dance track"
+	timesig := "4/4"
+
+	_, err = client.SoundGeneration(context.Background(), &pb.SoundGenerationRequest{
+		Text:          caption,
+		Caption:       &caption,
+		Dst:           outputFile,
+		Duration:      &duration,
+		Temperature:   &temperature,
+		Bpm:           &bpm,
+		Timesignature: &timesig,
+	})
+	if err != nil {
+		t.Fatalf("SoundGeneration failed: %v", err)
+	}
+
+	// Verify output file exists and has content
+	info, err := os.Stat(outputFile)
+	if os.IsNotExist(err) {
+		t.Fatal("Output audio file was not created")
+	}
+	if err != nil {
+		t.Fatalf("Failed to stat output file: %v", err)
+	}
+
+	t.Logf("Output file size: %d bytes", info.Size())
+
+	// WAV header is 44 bytes minimum; any real audio should be much larger
+	if info.Size() < 1000 {
+		t.Errorf("Output file too small (%d bytes), expected real audio data", info.Size())
+	}
+}

backend/go/acestep-cpp/cpp/goacestepcpp.cpp

@@ -0,0 +1,306 @@
+#include "goacestepcpp.h"
+#include "ggml-backend.h"
+
+#include "audio-io.h"
+#include "bpe.h"
+#include "cond-enc.h"
+#include "dit-sampler.h"
+#include "dit.h"
+#include "gguf-weights.h"
+#include "philox.h"
+#include "qwen3-enc.h"
+#include "qwen3-lm.h"
+#include "request.h"
+#include "vae.h"
+
+#include <cmath>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <random>
+#include <string>
+#include <vector>
+
+// Global model contexts (loaded once, reused across requests)
+static DiTGGML       g_dit       = {};
+static DiTGGMLConfig g_dit_cfg;
+static VAEGGML       g_vae       = {};
+static bool          g_dit_loaded = false;
+static bool          g_vae_loaded = false;
+static bool          g_is_turbo   = false;
+
+// Silence latent [15000, 64] — read once from DiT GGUF
+static std::vector<float> g_silence_full;
+
+// Paths for per-request loading (text encoder, tokenizer)
+static std::string g_text_enc_path;
+static std::string g_dit_path;
+static std::string g_lm_path;
+
+static void ggml_log_cb(enum ggml_log_level level, const char * log, void * data) {
+    const char * level_str;
+    if (!log)
+        return;
+    switch (level) {
+    case GGML_LOG_LEVEL_DEBUG:
+        level_str = "DEBUG";
+        break;
+    case GGML_LOG_LEVEL_INFO:
+        level_str = "INFO";
+        break;
+    case GGML_LOG_LEVEL_WARN:
+        level_str = "WARN";
+        break;
+    case GGML_LOG_LEVEL_ERROR:
+        level_str = "ERROR";
+        break;
+    default:
+        level_str = "?????";
+        break;
+    }
+    fprintf(stderr, "[%-5s] ", level_str);
+    fputs(log, stderr);
+    fflush(stderr);
+}
+
+int load_model(const char * lm_model_path, const char * text_encoder_path,
+               const char * dit_model_path, const char * vae_model_path) {
+    ggml_log_set(ggml_log_cb, nullptr);
+    ggml_backend_load_all();
+
+    g_lm_path       = lm_model_path;
+    g_text_enc_path = text_encoder_path;
+    g_dit_path      = dit_model_path;
+
+    // Load DiT model
+    fprintf(stderr, "[acestep-cpp] Loading DiT from %s\n", dit_model_path);
+    dit_ggml_init_backend(&g_dit);
+    if (!dit_ggml_load(&g_dit, dit_model_path, g_dit_cfg, nullptr, 0.0f)) {
+        fprintf(stderr, "[acestep-cpp] FATAL: failed to load DiT from %s\n", dit_model_path);
+        return 1;
+    }
+    g_dit_loaded = true;
+
+    // Read DiT GGUF metadata + silence_latent
+    {
+        GGUFModel gf = {};
+        if (gf_load(&gf, dit_model_path)) {
+            g_is_turbo           = gf_get_bool(gf, "acestep.is_turbo");
+            const void * sl_data = gf_get_data(gf, "silence_latent");
+            if (sl_data) {
+                g_silence_full.resize(15000 * 64);
+                memcpy(g_silence_full.data(), sl_data, 15000 * 64 * sizeof(float));
+                fprintf(stderr, "[acestep-cpp] silence_latent: [15000, 64] loaded\n");
+            } else {
+                fprintf(stderr, "[acestep-cpp] FATAL: silence_latent not found in %s\n", dit_model_path);
+                gf_close(&gf);
+                return 2;
+            }
+            gf_close(&gf);
+        } else {
+            fprintf(stderr, "[acestep-cpp] FATAL: cannot read GGUF metadata from %s\n", dit_model_path);
+            return 2;
+        }
+    }
+
+    // Load VAE model
+    fprintf(stderr, "[acestep-cpp] Loading VAE from %s\n", vae_model_path);
+    vae_ggml_load(&g_vae, vae_model_path);
+    g_vae_loaded = true;
+
+    fprintf(stderr, "[acestep-cpp] All models loaded successfully (turbo=%d)\n", g_is_turbo);
+    return 0;
+}
+
+int generate_music(const char * caption, const char * lyrics, int bpm,
+                   const char * keyscale, const char * timesignature,
+                   float duration, float temperature, bool instrumental,
+                   int seed, const char * dst, int threads) {
+    if (!g_dit_loaded || !g_vae_loaded) {
+        fprintf(stderr, "[acestep-cpp] ERROR: models not loaded\n");
+        return 1;
+    }
+
+    const int FRAMES_PER_SECOND = 25;
+
+    // Defaults
+    if (duration <= 0)
+        duration = 30.0f;
+    std::string cap_str    = caption ? caption : "";
+    std::string lyrics_str = (instrumental || !lyrics) ? "" : lyrics;
+    std::string ks_str     = keyscale ? keyscale : "N/A";
+    std::string ts_str     = timesignature ? timesignature : "4/4";
+    std::string lang_str   = "unknown";
+    char        bpm_str[16];
+    if (bpm > 0) {
+        snprintf(bpm_str, sizeof(bpm_str), "%d", bpm);
+    } else {
+        snprintf(bpm_str, sizeof(bpm_str), "N/A");
+    }
+
+    int   num_steps      = 8;
+    float guidance_scale = g_is_turbo ? 1.0f : 7.0f;
+    float shift          = 1.0f;
+
+    if (seed < 0) {
+        std::random_device rd;
+        seed = (int)(rd() & 0x7FFFFFFF);
+    }
+
+    // Compute T (latent frames at 25Hz)
+    int T = (int)(duration * FRAMES_PER_SECOND);
+    T     = ((T + g_dit_cfg.patch_size - 1) / g_dit_cfg.patch_size) * g_dit_cfg.patch_size;
+    int S = T / g_dit_cfg.patch_size;
+
+    if (T > 15000) {
+        fprintf(stderr, "[acestep-cpp] ERROR: T=%d exceeds max 15000\n", T);
+        return 2;
+    }
+
+    int Oc     = g_dit_cfg.out_channels;      // 64
+    int ctx_ch = g_dit_cfg.in_channels - Oc;  // 128
+
+    fprintf(stderr, "[acestep-cpp] T=%d, S=%d, duration=%.1fs, seed=%d\n", T, S, duration, seed);
+
+    // 1. Load BPE tokenizer from text encoder GGUF
+    BPETokenizer tok;
+    if (!load_bpe_from_gguf(&tok, g_text_enc_path.c_str())) {
+        fprintf(stderr, "[acestep-cpp] FATAL: failed to load BPE tokenizer\n");
+        return 3;
+    }
+
+    // 2. Build formatted prompts (matches dit-vae.cpp text2music template)
+    std::string instruction = "Fill the audio semantic mask based on the given conditions:";
+
+    char metas[512];
+    snprintf(metas, sizeof(metas),
+             "- bpm: %s\n- timesignature: %s\n- keyscale: %s\n- duration: %d seconds\n",
+             bpm_str, ts_str.c_str(), ks_str.c_str(), (int)duration);
+
+    std::string text_str  = std::string("# Instruction\n") + instruction + "\n\n" +
+                            "# Caption\n" + cap_str + "\n\n" +
+                            "# Metas\n" + metas + "<|endoftext|>\n";
+    std::string lyric_str = std::string("# Languages\n") + lang_str + "\n\n# Lyric\n" +
+                            lyrics_str + "<|endoftext|>";
+
+    // 3. Tokenize
+    auto text_ids  = bpe_encode(&tok, text_str.c_str(), true);
+    auto lyric_ids = bpe_encode(&tok, lyric_str.c_str(), true);
+    int  S_text    = (int)text_ids.size();
+    int  S_lyric   = (int)lyric_ids.size();
+
+    fprintf(stderr, "[acestep-cpp] caption: %d tokens, lyrics: %d tokens\n", S_text, S_lyric);
+
+    // 4. Text encoder forward
+    Qwen3GGML text_enc = {};
+    qwen3_init_backend(&text_enc);
+    if (!qwen3_load_text_encoder(&text_enc, g_text_enc_path.c_str())) {
+        fprintf(stderr, "[acestep-cpp] FATAL: failed to load text encoder\n");
+        return 4;
+    }
+
+    int                H_text = text_enc.cfg.hidden_size;  // 1024
+    std::vector<float> text_hidden(H_text * S_text);
+
+    qwen3_forward(&text_enc, text_ids.data(), S_text, text_hidden.data());
+    fprintf(stderr, "[acestep-cpp] TextEncoder forward done\n");
+
+    // 5. Lyric embedding
+    std::vector<float> lyric_embed(H_text * S_lyric);
+    qwen3_embed_lookup(&text_enc, lyric_ids.data(), S_lyric, lyric_embed.data());
+
+    // 6. Condition encoder
+    CondGGML cond = {};
+    cond_ggml_init_backend(&cond);
+    if (!cond_ggml_load(&cond, g_dit_path.c_str())) {
+        fprintf(stderr, "[acestep-cpp] FATAL: failed to load condition encoder\n");
+        qwen3_free(&text_enc);
+        return 5;
+    }
+
+    const int          S_ref = 750;
+    std::vector<float> silence_feats(S_ref * 64);
+    memcpy(silence_feats.data(), g_silence_full.data(), S_ref * 64 * sizeof(float));
+
+    int                enc_S = 0;
+    std::vector<float> enc_hidden;
+    cond_ggml_forward(&cond, text_hidden.data(), S_text, lyric_embed.data(), S_lyric,
+                      silence_feats.data(), S_ref, enc_hidden, &enc_S);
+    fprintf(stderr, "[acestep-cpp] ConditionEncoder done, enc_S=%d\n", enc_S);
+
+    qwen3_free(&text_enc);
+    cond_ggml_free(&cond);
+
+    // 7. Build context [T, ctx_ch] = silence[64] + mask[64]
+    std::vector<float> context(T * ctx_ch);
+    for (int t = 0; t < T; t++) {
+        const float * src = g_silence_full.data() + t * Oc;
+        for (int c = 0; c < Oc; c++) {
+            context[t * ctx_ch + c] = src[c];
+        }
+        for (int c = 0; c < Oc; c++) {
+            context[t * ctx_ch + Oc + c] = 1.0f;
+        }
+    }
+
+    // 8. Build schedule
+    std::vector<float> schedule(num_steps);
+    for (int i = 0; i < num_steps; i++) {
+        float t     = 1.0f - (float)i / (float)num_steps;
+        schedule[i] = shift * t / (1.0f + (shift - 1.0f) * t);
+    }
+
+    // 9. Generate noise (Philox)
+    std::vector<float> noise(Oc * T);
+    philox_randn((long long)seed, noise.data(), Oc * T, true);
+
+    // 10. DiT generate
+    std::vector<float> output(Oc * T);
+    fprintf(stderr, "[acestep-cpp] DiT generate: T=%d, steps=%d, guidance=%.1f\n", T, num_steps, guidance_scale);
+
+    dit_ggml_generate(&g_dit, noise.data(), context.data(), enc_hidden.data(), enc_S,
+                      T, 1, num_steps, schedule.data(), output.data(), guidance_scale,
+                      nullptr, nullptr, -1);
+    fprintf(stderr, "[acestep-cpp] DiT generation done\n");
+
+    // 11. VAE decode
+    int                T_audio_max = T * 1920;
+    std::vector<float> audio(2 * T_audio_max);
+
+    int T_audio = vae_ggml_decode_tiled(&g_vae, output.data(), T, audio.data(), T_audio_max, 256, 64);
+    if (T_audio < 0) {
+        fprintf(stderr, "[acestep-cpp] ERROR: VAE decode failed\n");
+        return 6;
+    }
+    fprintf(stderr, "[acestep-cpp] VAE decode done: %d samples (%.2fs @ 48kHz)\n", T_audio,
+            (float)T_audio / 48000.0f);
+
+    // 12. Peak normalization to -1.0 dB
+    {
+        float peak      = 0.0f;
+        int   n_samples = 2 * T_audio;
+        for (int i = 0; i < n_samples; i++) {
+            float a = audio[i] < 0 ? -audio[i] : audio[i];
+            if (a > peak) {
+                peak = a;
+            }
+        }
+        if (peak > 1e-6f) {
+            const float target_amp = powf(10.0f, -1.0f / 20.0f);
+            float       gain       = target_amp / peak;
+            for (int i = 0; i < n_samples; i++) {
+                audio[i] *= gain;
+            }
+        }
+    }
+
+    // 13. Write WAV output
+    if (!audio_write_wav(dst, audio.data(), T_audio, 48000)) {
+        fprintf(stderr, "[acestep-cpp] ERROR: failed to write %s\n", dst);
+        return 7;
+    }
+
+    fprintf(stderr, "[acestep-cpp] Wrote %s: %d samples (%.2fs @ 48kHz stereo)\n",
+            dst, T_audio, (float)T_audio / 48000.0f);
+    return 0;
+}

backend/go/acestep-cpp/cpp/goacestepcpp.h

@@ -0,0 +1,11 @@
+#include <cstddef>
+#include <cstdint>
+
+extern "C" {
+int load_model(const char *lm_model_path, const char *text_encoder_path,
+               const char *dit_model_path, const char *vae_model_path);
+int generate_music(const char *caption, const char *lyrics, int bpm,
+                   const char *keyscale, const char *timesignature,
+                   float duration, float temperature, bool instrumental,
+                   int seed, const char *dst, int threads);
+}

backend/go/acestep-cpp/goacestepcpp.go

@@ -0,0 +1,109 @@
+package main
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"strings"
+
+	"github.com/mudler/LocalAI/pkg/grpc/base"
+	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
+)
+
+var (
+	CppLoadModel     func(lmModelPath, textEncoderPath, ditModelPath, vaeModelPath string) int
+	CppGenerateMusic func(caption, lyrics string, bpm int, keyscale, timesignature string, duration, temperature float32, instrumental bool, seed int, dst string, threads int) int
+)
+
+type AceStepCpp struct {
+	base.SingleThread
+}
+
+func (a *AceStepCpp) Load(opts *pb.ModelOptions) error {
+	// ModelFile is the LM model path
+	lmModel := opts.ModelFile
+
+	// Get the base directory from ModelFile for resolving relative paths
+	baseDir := opts.ModelPath
+
+	var textEncoderModel, ditModel, vaeModel string
+
+	for _, oo := range opts.Options {
+		key, value, found := strings.Cut(oo, ":")
+		if !found {
+			fmt.Fprintf(os.Stderr, "Unrecognized option: %v\n", oo)
+			continue
+		}
+		switch key {
+		case "text_encoder_model":
+			textEncoderModel = value
+		case "dit_model":
+			ditModel = value
+		case "vae_model":
+			vaeModel = value
+		default:
+			fmt.Fprintf(os.Stderr, "Unrecognized option: %v\n", oo)
+		}
+	}
+
+	if textEncoderModel == "" {
+		return fmt.Errorf("text_encoder_model option is required")
+	}
+	if ditModel == "" {
+		return fmt.Errorf("dit_model option is required")
+	}
+	if vaeModel == "" {
+		return fmt.Errorf("vae_model option is required")
+	}
+
+	// Resolve relative paths to the base directory
+	// If the path doesn't start with "/" it's relative
+	if !filepath.IsAbs(textEncoderModel) {
+		textEncoderModel = filepath.Join(baseDir, textEncoderModel)
+	}
+	if !filepath.IsAbs(ditModel) {
+		ditModel = filepath.Join(baseDir, ditModel)
+	}
+	if !filepath.IsAbs(vaeModel) {
+		vaeModel = filepath.Join(baseDir, vaeModel)
+	}
+
+	// Also resolve the lmModel if it's relative
+	if !filepath.IsAbs(lmModel) {
+		lmModel = filepath.Join(baseDir, lmModel)
+	}
+
+	fmt.Fprintf(os.Stderr, "[acestep-cpp] Resolved paths:\n")
+	fmt.Fprintf(os.Stderr, "  LM Model: %s\n", lmModel)
+	fmt.Fprintf(os.Stderr, "  Text Encoder: %s\n", textEncoderModel)
+	fmt.Fprintf(os.Stderr, "  DiT Model: %s\n", ditModel)
+	fmt.Fprintf(os.Stderr, "  VAE Model: %s\n", vaeModel)
+
+	if ret := CppLoadModel(lmModel, textEncoderModel, ditModel, vaeModel); ret != 0 {
+		return fmt.Errorf("failed to load acestep models (error code: %d)", ret)
+	}
+
+	return nil
+}
+
+func (a *AceStepCpp) SoundGeneration(req *pb.SoundGenerationRequest) error {
+	caption := req.GetCaption()
+	if caption == "" {
+		caption = req.GetText()
+	}
+	lyrics := req.GetLyrics()
+	bpm := int(req.GetBpm())
+	keyscale := req.GetKeyscale()
+	timesignature := req.GetTimesignature()
+	duration := req.GetDuration()
+	temperature := req.GetTemperature()
+	instrumental := req.GetInstrumental()
+	seed := 42
+	threads := 4
+
+	if ret := CppGenerateMusic(caption, lyrics, bpm, keyscale, timesignature, duration, temperature, instrumental, seed, req.GetDst(), threads); ret != 0 {
+		return fmt.Errorf("failed to generate music (error code: %d)", ret)
+	}
+
+	return nil
+}

backend/go/acestep-cpp/main.go

@@ -0,0 +1,47 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+import (
+	"flag"
+	"os"
+
+	"github.com/ebitengine/purego"
+	grpc "github.com/mudler/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+type LibFuncs struct {
+	FuncPtr any
+	Name    string
+}
+
+func main() {
+	// Get library name from environment variable, default to fallback
+	libName := os.Getenv("ACESTEP_LIBRARY")
+	if libName == "" {
+		libName = "./libgoacestepcpp-fallback.so"
+	}
+
+	gosd, err := purego.Dlopen(libName, purego.RTLD_NOW|purego.RTLD_GLOBAL)
+	if err != nil {
+		panic(err)
+	}
+
+	libFuncs := []LibFuncs{
+		{&CppLoadModel, "load_model"},
+		{&CppGenerateMusic, "generate_music"},
+	}
+
+	for _, lf := range libFuncs {
+		purego.RegisterLibFunc(lf.FuncPtr, gosd, lf.Name)
+	}
+
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &AceStepCpp{}); err != nil {
+		panic(err)
+	}
+}

backend/go/acestep-cpp/package.sh

@@ -0,0 +1,65 @@
+#!/bin/bash
+
+# Script to copy the appropriate libraries based on architecture
+# This script is used in the final stage of the Dockerfile
+
+set -e
+
+CURDIR=$(dirname "$(realpath $0)")
+REPO_ROOT="${CURDIR}/../../.."
+
+# Create lib directory
+mkdir -p $CURDIR/package/lib
+
+cp -avf $CURDIR/acestep-cpp $CURDIR/package/
+cp -fv $CURDIR/libgoacestepcpp-*.so $CURDIR/package/
+cp -fv $CURDIR/run.sh $CURDIR/package/
+
+# Detect architecture and copy appropriate libraries
+if [ -f "/lib64/ld-linux-x86-64.so.2" ]; then
+    # x86_64 architecture
+    echo "Detected x86_64 architecture, copying x86_64 libraries..."
+    cp -arfLv /lib64/ld-linux-x86-64.so.2 $CURDIR/package/lib/ld.so
+    cp -arfLv /lib/x86_64-linux-gnu/libc.so.6 $CURDIR/package/lib/libc.so.6
+    cp -arfLv /lib/x86_64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
+    cp -arfLv /lib/x86_64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
+    cp -arfLv /lib/x86_64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
+    cp -arfLv /lib/x86_64-linux-gnu/libdl.so.2 $CURDIR/package/lib/libdl.so.2
+    cp -arfLv /lib/x86_64-linux-gnu/librt.so.1 $CURDIR/package/lib/librt.so.1
+    cp -arfLv /lib/x86_64-linux-gnu/libpthread.so.0 $CURDIR/package/lib/libpthread.so.0
+elif [ -f "/lib/ld-linux-aarch64.so.1" ]; then
+    # ARM64 architecture
+    echo "Detected ARM64 architecture, copying ARM64 libraries..."
+    cp -arfLv /lib/ld-linux-aarch64.so.1 $CURDIR/package/lib/ld.so
+    cp -arfLv /lib/aarch64-linux-gnu/libc.so.6 $CURDIR/package/lib/libc.so.6
+    cp -arfLv /lib/aarch64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
+    cp -arfLv /lib/aarch64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
+    cp -arfLv /lib/aarch64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
+    cp -arfLv /lib/aarch64-linux-gnu/libdl.so.2 $CURDIR/package/lib/libdl.so.2
+    cp -arfLv /lib/aarch64-linux-gnu/librt.so.1 $CURDIR/package/lib/librt.so.1
+    cp -arfLv /lib/aarch64-linux-gnu/libpthread.so.0 $CURDIR/package/lib/libpthread.so.0
+elif [ $(uname -s) = "Darwin" ]; then
+    echo "Detected Darwin"
+else
+    echo "Error: Could not detect architecture"
+    exit 1
+fi
+
+# Package GPU libraries based on BUILD_TYPE
+# The GPU library packaging script will detect BUILD_TYPE and copy appropriate GPU libraries
+GPU_LIB_SCRIPT="${REPO_ROOT}/scripts/build/package-gpu-libs.sh"
+if [ -f "$GPU_LIB_SCRIPT" ]; then
+    echo "Packaging GPU libraries for BUILD_TYPE=${BUILD_TYPE:-cpu}..."
+    source "$GPU_LIB_SCRIPT" "$CURDIR/package/lib"
+    package_gpu_libs
+fi
+
+echo "Packaging completed successfully"
+ls -liah $CURDIR/package/
+ls -liah $CURDIR/package/lib/

backend/go/acestep-cpp/run.sh

@@ -0,0 +1,52 @@
+#!/bin/bash
+set -ex
+
+# Get the absolute current dir where the script is located
+CURDIR=$(dirname "$(realpath $0)")
+
+cd /
+
+echo "CPU info:"
+if [ "$(uname)" != "Darwin" ]; then
+	grep -e "model\sname" /proc/cpuinfo | head -1
+	grep -e "flags" /proc/cpuinfo | head -1
+fi
+
+LIBRARY="$CURDIR/libgoacestepcpp-fallback.so"
+
+if [ "$(uname)" != "Darwin" ]; then
+	if grep -q -e "\savx\s" /proc/cpuinfo ; then
+		echo "CPU:    AVX    found OK"
+		if [ -e $CURDIR/libgoacestepcpp-avx.so ]; then
+			LIBRARY="$CURDIR/libgoacestepcpp-avx.so"
+		fi
+	fi
+
+	if grep -q -e "\savx2\s" /proc/cpuinfo ; then
+		echo "CPU:    AVX2   found OK"
+		if [ -e $CURDIR/libgoacestepcpp-avx2.so ]; then
+			LIBRARY="$CURDIR/libgoacestepcpp-avx2.so"
+		fi
+	fi
+
+	# Check avx 512
+	if grep -q -e "\savx512f\s" /proc/cpuinfo ; then
+		echo "CPU:    AVX512F found OK"
+		if [ -e $CURDIR/libgoacestepcpp-avx512.so ]; then
+			LIBRARY="$CURDIR/libgoacestepcpp-avx512.so"
+		fi
+	fi
+fi
+
+export LD_LIBRARY_PATH=$CURDIR/lib:$LD_LIBRARY_PATH
+export ACESTEP_LIBRARY=$LIBRARY
+
+# If there is a lib/ld.so, use it
+if [ -f $CURDIR/lib/ld.so ]; then
+	echo "Using lib/ld.so"
+	echo "Using library: $LIBRARY"
+	exec $CURDIR/lib/ld.so $CURDIR/acestep-cpp "$@"
+fi
+
+echo "Using library: $LIBRARY"
+exec $CURDIR/acestep-cpp "$@"

backend/go/acestep-cpp/test.sh

@@ -0,0 +1,54 @@
+#!/bin/bash
+set -e
+
+CURDIR=$(dirname "$(realpath $0)")
+
+echo "Running acestep-cpp backend tests..."
+
+# The test requires:
+#   - ACESTEP_MODEL_DIR: path to directory containing GGUF model files
+#   - ACESTEP_BINARY: path to the acestep-cpp binary (defaults to ./acestep-cpp)
+#
+# Tests that require the model will be skipped if ACESTEP_MODEL_DIR is not set
+# or the directory does not contain the required model files.
+
+cd "$CURDIR"
+
+# Only auto-download models when ACESTEP_MODEL_DIR is not explicitly set
+if [ -z "$ACESTEP_MODEL_DIR" ]; then
+    export ACESTEP_MODEL_DIR="./acestep-models"
+
+    if [ ! -d "$ACESTEP_MODEL_DIR" ]; then
+        echo "Creating acestep-models directory for tests..."
+        mkdir -p "$ACESTEP_MODEL_DIR"
+        REPO_ID="Serveurperso/ACE-Step-1.5-GGUF"
+        echo "Repository: ${REPO_ID}"
+        echo ""
+
+        # Files to download (smallest quantizations for testing)
+        FILES=(
+            "acestep-5Hz-lm-0.6B-Q8_0.gguf"
+            "Qwen3-Embedding-0.6B-Q8_0.gguf"
+            "acestep-v15-turbo-Q8_0.gguf"
+            "vae-BF16.gguf"
+        )
+
+        BASE_URL="https://huggingface.co/${REPO_ID}/resolve/main"
+
+        for file in "${FILES[@]}"; do
+            dest="${ACESTEP_MODEL_DIR}/${file}"
+            if [ -f "${dest}" ]; then
+                echo "  [skip] ${file} (already exists)"
+            else
+                echo "  [download] ${file}..."
+                curl -L -o "${dest}" "${BASE_URL}/${file}" --progress-bar
+                echo "  [done] ${file}"
+            fi
+        done
+    fi
+fi
+
+# Run Go tests
+go test -v -timeout 600s .
+
+echo "All acestep-cpp tests passed."

backend/go/huggingface/Makefile

@@ -1,12 +0,0 @@
-GOCMD=go
-
-huggingface:
-	CGO_ENABLED=0 $(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o huggingface ./
-
-package:
-	bash package.sh
-
-build: huggingface package
-
-clean:
-	rm -f huggingface

backend/go/huggingface/langchain.go

@@ -1,64 +0,0 @@
-package main
-
-// This is a wrapper to statisfy the GRPC service interface
-// It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
-import (
-	"fmt"
-	"os"
-
-	"github.com/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-	"github.com/mudler/LocalAI/pkg/langchain"
-)
-
-type LLM struct {
-	base.Base
-
-	langchain *langchain.HuggingFace
-	model     string
-}
-
-func (llm *LLM) Load(opts *pb.ModelOptions) error {
-	var err error
-	hfToken := os.Getenv("HUGGINGFACEHUB_API_TOKEN")
-	if hfToken == "" {
-		return fmt.Errorf("no huggingface token provided")
-	}
-	llm.langchain, err = langchain.NewHuggingFace(opts.Model, hfToken)
-	llm.model = opts.Model
-	return err
-}
-
-func (llm *LLM) Predict(opts *pb.PredictOptions) (string, error) {
-	o := []langchain.PredictOption{
-		langchain.SetModel(llm.model),
-		langchain.SetMaxTokens(int(opts.Tokens)),
-		langchain.SetTemperature(float64(opts.Temperature)),
-		langchain.SetStopWords(opts.StopPrompts),
-	}
-	pred, err := llm.langchain.PredictHuggingFace(opts.Prompt, o...)
-	if err != nil {
-		return "", err
-	}
-	return pred.Completion, nil
-}
-
-func (llm *LLM) PredictStream(opts *pb.PredictOptions, results chan string) error {
-	o := []langchain.PredictOption{
-		langchain.SetModel(llm.model),
-		langchain.SetMaxTokens(int(opts.Tokens)),
-		langchain.SetTemperature(float64(opts.Temperature)),
-		langchain.SetStopWords(opts.StopPrompts),
-	}
-	go func() {
-		res, err := llm.langchain.PredictHuggingFace(opts.Prompt, o...)
-
-		if err != nil {
-			fmt.Println("err: ", err)
-		}
-		results <- res.Completion
-		close(results)
-	}()
-
-	return nil
-}

backend/go/huggingface/main.go

@@ -1,21 +0,0 @@
-package main
-
-// Note: this is started internally by LocalAI and a server is allocated for each model
-
-import (
-	"flag"
-
-	grpc "github.com/mudler/LocalAI/pkg/grpc"
-)
-
-var (
-	addr = flag.String("addr", "localhost:50051", "the address to connect to")
-)
-
-func main() {
-	flag.Parse()
-
-	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
-		panic(err)
-	}
-}

backend/go/huggingface/package.sh

@@ -1,12 +0,0 @@
-#!/bin/bash
-
-# Script to copy the appropriate libraries based on architecture
-# This script is used in the final stage of the Dockerfile
-
-set -e
-
-CURDIR=$(dirname "$(realpath $0)")
-
-mkdir -p $CURDIR/package
-cp -avf $CURDIR/huggingface $CURDIR/package/
-cp -rfv $CURDIR/run.sh $CURDIR/package/

backend/go/huggingface/run.sh

@@ -1,6 +0,0 @@
-#!/bin/bash
-set -ex
-
-CURDIR=$(dirname "$(realpath $0)")
-
-exec $CURDIR/huggingface "$@"

backend/go/llm/llama/llama.go

@@ -18,6 +18,19 @@ type LLM struct {
 	draftModel *llama.LLama
 }
 
+// Free releases GPU resources and frees the llama model
+// This should be called when the model is being unloaded to properly release VRAM
+func (llm *LLM) Free() error {
+	if llm.llama != nil {
+		llm.llama.Free()
+		llm.llama = nil
+	}
+	if llm.draftModel != nil {
+		llm.draftModel.Free()
+		llm.draftModel = nil
+	}
+	return nil
+}
 func (llm *LLM) Load(opts *pb.ModelOptions) error {
 	ropeFreqBase := float32(10000)
 	ropeFreqScale := float32(1)

backend/go/local-store/debug.go

@@ -1,5 +1,4 @@
 //go:build debug
-// +build debug
 
 package main
 

backend/go/local-store/production.go

@@ -1,5 +1,4 @@
 //go:build !debug
-// +build !debug
 
 package main
 

backend/go/local-store/store.go

@@ -332,7 +332,7 @@ func normalizedCosineSimilarity(k1, k2 []float32) float32 {
 	assert(len(k1) == len(k2), fmt.Sprintf("normalizedCosineSimilarity: len(k1) = %d, len(k2) = %d", len(k1), len(k2)))
 
 	var dot float32
-	for i := 0; i < len(k1); i++ {
+	for i := range len(k1) {
 		dot += k1[i] * k2[i]
 	}
 
@@ -419,7 +419,7 @@ func cosineSimilarity(k1, k2 []float32, mag1 float64) float32 {
 	assert(len(k1) == len(k2), fmt.Sprintf("cosineSimilarity: len(k1) = %d, len(k2) = %d", len(k1), len(k2)))
 
 	var dot, mag2 float64
-	for i := 0; i < len(k1); i++ {
+	for i := range len(k1) {
 		dot += float64(k1[i] * k2[i])
 		mag2 += float64(k2[i] * k2[i])
 	}

backend/go/opus/Makefile

@@ -0,0 +1,19 @@
+GOCMD?=go
+GO_TAGS?=
+
+OPUS_CFLAGS := $(shell pkg-config --cflags opus)
+OPUS_LIBS := $(shell pkg-config --libs opus)
+
+libopusshim.so: csrc/opus_shim.c
+	$(CC) -shared -fPIC -o $@ $< $(OPUS_CFLAGS) $(OPUS_LIBS)
+
+opus: libopusshim.so
+	$(GOCMD) build -tags "$(GO_TAGS)" -o opus ./
+
+package: opus
+	bash package.sh
+
+build: package
+
+clean:
+	rm -f opus libopusshim.so

backend/go/opus/codec.go

@@ -0,0 +1,256 @@
+package main
+
+import (
+	"errors"
+	"fmt"
+	"os"
+	"path/filepath"
+	"runtime"
+	"sync"
+
+	"github.com/ebitengine/purego"
+)
+
+const (
+	ApplicationVoIP               = 2048
+	ApplicationAudio              = 2049
+	ApplicationRestrictedLowDelay = 2051
+)
+
+var (
+	initOnce sync.Once
+	initErr  error
+
+	opusLib uintptr
+	shimLib uintptr
+
+	// libopus functions
+	cEncoderCreate  func(fs int32, channels int32, application int32, errPtr *int32) uintptr
+	cEncode         func(st uintptr, pcm *int16, frameSize int32, data *byte, maxBytes int32) int32
+	cEncoderDestroy func(st uintptr)
+
+	cDecoderCreate  func(fs int32, channels int32, errPtr *int32) uintptr
+	cDecode         func(st uintptr, data *byte, dataLen int32, pcm *int16, frameSize int32, decodeFec int32) int32
+	cDecoderDestroy func(st uintptr)
+
+	// shim functions (non-variadic wrappers for opus_encoder_ctl)
+	cSetBitrate    func(st uintptr, bitrate int32) int32
+	cSetComplexity func(st uintptr, complexity int32) int32
+)
+
+func loadLib(names []string) (uintptr, error) {
+	var firstErr error
+	for _, name := range names {
+		h, err := purego.Dlopen(name, purego.RTLD_NOW|purego.RTLD_GLOBAL)
+		if err == nil {
+			return h, nil
+		}
+		if firstErr == nil {
+			firstErr = err
+		}
+	}
+	return 0, firstErr
+}
+
+func ensureInit() error {
+	initOnce.Do(func() {
+		initErr = doInit()
+	})
+	return initErr
+}
+
+const shimHint = "ensure libopus-dev is installed and rebuild, or set OPUS_LIBRARY / OPUS_SHIM_LIBRARY env vars"
+
+func doInit() error {
+	opusNames := opusSearchPaths()
+	var err error
+	opusLib, err = loadLib(opusNames)
+	if err != nil {
+		return fmt.Errorf("opus: failed to load libopus (%s): %w", shimHint, err)
+	}
+
+	purego.RegisterLibFunc(&cEncoderCreate, opusLib, "opus_encoder_create")
+	purego.RegisterLibFunc(&cEncode, opusLib, "opus_encode")
+	purego.RegisterLibFunc(&cEncoderDestroy, opusLib, "opus_encoder_destroy")
+	purego.RegisterLibFunc(&cDecoderCreate, opusLib, "opus_decoder_create")
+	purego.RegisterLibFunc(&cDecode, opusLib, "opus_decode")
+	purego.RegisterLibFunc(&cDecoderDestroy, opusLib, "opus_decoder_destroy")
+
+	shimNames := shimSearchPaths()
+	shimLib, err = loadLib(shimNames)
+	if err != nil {
+		return fmt.Errorf("opus: failed to load libopusshim (%s): %w", shimHint, err)
+	}
+
+	purego.RegisterLibFunc(&cSetBitrate, shimLib, "opus_shim_encoder_set_bitrate")
+	purego.RegisterLibFunc(&cSetComplexity, shimLib, "opus_shim_encoder_set_complexity")
+
+	return nil
+}
+
+func opusSearchPaths() []string {
+	var paths []string
+
+	if env := os.Getenv("OPUS_LIBRARY"); env != "" {
+		paths = append(paths, env)
+	}
+
+	if exe, err := os.Executable(); err == nil {
+		dir := filepath.Dir(exe)
+		paths = append(paths, filepath.Join(dir, "libopus.so.0"), filepath.Join(dir, "libopus.so"))
+		if runtime.GOOS == "darwin" {
+			paths = append(paths, filepath.Join(dir, "libopus.dylib"))
+		}
+	}
+
+	paths = append(paths, "libopus.so.0", "libopus.so", "libopus.dylib", "opus.dll")
+
+	if runtime.GOOS == "darwin" {
+		paths = append(paths,
+			"/opt/homebrew/lib/libopus.dylib",
+			"/usr/local/lib/libopus.dylib",
+		)
+	}
+
+	return paths
+}
+
+func shimSearchPaths() []string {
+	var paths []string
+
+	if env := os.Getenv("OPUS_SHIM_LIBRARY"); env != "" {
+		paths = append(paths, env)
+	}
+
+	if exe, err := os.Executable(); err == nil {
+		dir := filepath.Dir(exe)
+		paths = append(paths, filepath.Join(dir, "libopusshim.so"))
+		if runtime.GOOS == "darwin" {
+			paths = append(paths, filepath.Join(dir, "libopusshim.dylib"))
+		}
+	}
+
+	paths = append(paths, "./libopusshim.so", "libopusshim.so")
+	if runtime.GOOS == "darwin" {
+		paths = append(paths, "./libopusshim.dylib", "libopusshim.dylib")
+	}
+	return paths
+}
+
+// Encoder wraps a libopus OpusEncoder via purego.
+type Encoder struct {
+	st uintptr
+}
+
+func NewEncoder(sampleRate, channels, application int) (*Encoder, error) {
+	if err := ensureInit(); err != nil {
+		return nil, err
+	}
+
+	var opusErr int32
+	st := cEncoderCreate(int32(sampleRate), int32(channels), int32(application), &opusErr)
+	if opusErr != 0 || st == 0 {
+		return nil, fmt.Errorf("opus_encoder_create failed: error %d", opusErr)
+	}
+	return &Encoder{st: st}, nil
+}
+
+// Encode encodes a frame of PCM int16 samples. It returns the number of bytes
+// written to out, or a negative error code.
+func (e *Encoder) Encode(pcm []int16, frameSize int, out []byte) (int, error) {
+	if len(pcm) == 0 || len(out) == 0 {
+		return 0, errors.New("opus encode: empty input or output buffer")
+	}
+	n := cEncode(e.st, &pcm[0], int32(frameSize), &out[0], int32(len(out)))
+	if n < 0 {
+		return 0, fmt.Errorf("opus_encode failed: error %d", n)
+	}
+	return int(n), nil
+}
+
+func (e *Encoder) SetBitrate(bitrate int) error {
+	if ret := cSetBitrate(e.st, int32(bitrate)); ret != 0 {
+		return fmt.Errorf("opus set bitrate: error %d", ret)
+	}
+	return nil
+}
+
+func (e *Encoder) SetComplexity(complexity int) error {
+	if ret := cSetComplexity(e.st, int32(complexity)); ret != 0 {
+		return fmt.Errorf("opus set complexity: error %d", ret)
+	}
+	return nil
+}
+
+func (e *Encoder) Close() {
+	if e.st != 0 {
+		cEncoderDestroy(e.st)
+		e.st = 0
+	}
+}
+
+// Decoder wraps a libopus OpusDecoder via purego.
+type Decoder struct {
+	st uintptr
+}
+
+func NewDecoder(sampleRate, channels int) (*Decoder, error) {
+	if err := ensureInit(); err != nil {
+		return nil, err
+	}
+
+	var opusErr int32
+	st := cDecoderCreate(int32(sampleRate), int32(channels), &opusErr)
+	if opusErr != 0 || st == 0 {
+		return nil, fmt.Errorf("opus_decoder_create failed: error %d", opusErr)
+	}
+	return &Decoder{st: st}, nil
+}
+
+// Decode decodes an Opus packet into pcm. frameSize is the max number of
+// samples per channel that pcm can hold. Returns the number of decoded samples
+// per channel.
+func (d *Decoder) Decode(data []byte, pcm []int16, frameSize int, fec bool) (int, error) {
+	if len(pcm) == 0 {
+		return 0, errors.New("opus decode: empty output buffer")
+	}
+
+	var dataPtr *byte
+	var dataLen int32
+	if len(data) > 0 {
+		dataPtr = &data[0]
+		dataLen = int32(len(data))
+	}
+
+	decodeFec := int32(0)
+	if fec {
+		decodeFec = 1
+	}
+
+	n := cDecode(d.st, dataPtr, dataLen, &pcm[0], int32(frameSize), decodeFec)
+	if n < 0 {
+		return 0, fmt.Errorf("opus_decode failed: error %d", n)
+	}
+	return int(n), nil
+}
+
+func (d *Decoder) Close() {
+	if d.st != 0 {
+		cDecoderDestroy(d.st)
+		d.st = 0
+	}
+}
+
+// Init eagerly loads the opus libraries, returning any error.
+// Calling this is optional; the libraries are loaded lazily on first use.
+func Init() error {
+	return ensureInit()
+}
+
+// Reset allows re-initialization (for testing).
+func Reset() {
+	initOnce = sync.Once{}
+	initErr = nil
+	opusLib = 0
+	shimLib = 0
+}