[pull] master from mudler:master

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

@@ -52,6 +52,8 @@ jobs:
         run: |
           sudo apt-get update
           sudo apt-get install -y build-essential libopus-dev
+      - name: Run stale chunk recovery tests
+        run: PATH="$PATH:$HOME/go/bin" make test-ui-stale-chunk
       # Builds an instrumented UI bundle, runs the Playwright specs, and fails
       # if line coverage regressed beyond the jitter tolerance (the gate is
       # in `make test-ui-coverage-check`). PLAYWRIGHT_CHROMIUM_PATH is unset

Makefile

@@ -103,7 +103,7 @@ COVERAGE_E2E_LABELS?=!real-models
 COVERAGE_EXCLUDE_RE?=grpc/proto/.*[.]pb[.]go
 
 
-.PHONY: all test test-coverage test-coverage-baseline test-coverage-check test-backend-cpp test-build-scripts test-ui test-ui-coverage-baseline test-ui-coverage-check build vendor lint lint-all
+.PHONY: all test test-coverage test-coverage-baseline test-coverage-check test-backend-cpp test-build-scripts test-ui test-ui-stale-chunk test-ui-coverage-baseline test-ui-coverage-check build vendor lint lint-all
 
 all: help
 
@@ -1517,6 +1517,13 @@ test-ui: build-mock-backend protogen-go
 	$(GOCMD) build -o tests/e2e-ui/ui-test-server ./tests/e2e-ui
 	cd core/http/react-ui && sh $(CURDIR)/scripts/ensure-playwright-browser.sh && bunx playwright test $(PLAYWRIGHT_WORKERS_FLAG)
 
+## The stale-chunk specs need the production code-split bundle. The V8 coverage
+## bundle below inlines dynamic imports to keep every page in its denominator.
+test-ui-stale-chunk: build-mock-backend protogen-go
+	cd core/http/react-ui && bun install && bun run build
+	$(GOCMD) build -o tests/e2e-ui/ui-test-server ./tests/e2e-ui
+	cd core/http/react-ui && sh $(CURDIR)/scripts/ensure-playwright-browser.sh && bunx playwright test --grep @production-chunks --workers=1
+
 ## React UI code coverage from the Playwright e2e suite. Builds a
 ## NON-instrumented bundle with source maps (COVERAGE_V8=true), re-embeds it
 ## into the ui-test-server (the dist is //go:embed'ed at compile time), runs the
@@ -1532,7 +1539,7 @@ test-ui-coverage: build-mock-backend protogen-go
 	$(GOCMD) build -o tests/e2e-ui/ui-test-server ./tests/e2e-ui && \
 	( cd core/http/react-ui && rm -rf .nyc_output coverage && \
 	    sh $(CURDIR)/scripts/ensure-playwright-browser.sh && \
-	    PW_V8_COVERAGE=1 bunx playwright test $(PLAYWRIGHT_WORKERS_FLAG) && bun run coverage:report )
+	    PW_V8_COVERAGE=1 bunx playwright test --grep-invert @production-chunks $(PLAYWRIGHT_WORKERS_FLAG) && bun run coverage:report )
 
 ## UI coverage baseline (committed) and the strict gate that compares against
 ## it — the React mirror of test-coverage-baseline / test-coverage-check.

backend/cpp/audio-cpp/Makefile

@@ -9,7 +9,7 @@
 # recipe is a make target (not a prepare.sh) so 'make purge && make' is a clean
 # rebuild and so the bump bot can see the pin.
 
-AUDIO_CPP_VERSION?=3407559d7c83b543bd60ca8fa2b630374e71083f
+AUDIO_CPP_VERSION?=92816fc5c48d33a58171b49678ccf884d70142fb
 AUDIO_CPP_REPO?=https://github.com/0xShug0/audio.cpp
 
 CURRENT_MAKEFILE_DIR := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))

backend/cpp/llama-cpp/Makefile

@@ -1,5 +1,5 @@
 
-LLAMA_VERSION?=0021a77de0a8966059dc94548fb3b96654e0bb12
+LLAMA_VERSION?=60addddf3c567c43ec3caf70fc953fba3572d96f
 LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
 
 CMAKE_ARGS?=

backend/go/crispasr/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # CrispASR version (release tag)
 CRISPASR_REPO?=https://github.com/CrispStrobe/CrispASR
-CRISPASR_VERSION?=007fda47128aaba7e2ce2279b0947a1c13800d68
+CRISPASR_VERSION?=a117d433855ecc36b0cb20bd1abddd852902ed13
 SO_TARGET?=libgocrispasr.so
 
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF

backend/go/parakeet-cpp/Makefile

@@ -1,6 +1,6 @@
 # parakeet-cpp backend Makefile.
 #
-# Upstream pin lives below as PARAKEET_VERSION?=1bfbebfaaf493866f49597cd3b7901959d395c60
+# Upstream pin lives below as PARAKEET_VERSION?=e75de9b6b9b688fd293aa22f7e27aa724ea286f8
 # (.github/bump_deps.sh) can find and update it - matches the
 # whisper.cpp / ds4 / vibevoice-cpp convention.
 #
@@ -15,7 +15,7 @@
 # That's what the L0 smoke test uses. The default target below does the
 # proper clone-at-pin + cmake build so CI doesn't need a side-checkout.
 
-PARAKEET_VERSION?=1bfbebfaaf493866f49597cd3b7901959d395c60
+PARAKEET_VERSION?=e75de9b6b9b688fd293aa22f7e27aa724ea286f8
 PARAKEET_REPO?=https://github.com/mudler/parakeet.cpp
 
 GOCMD?=go

backend/go/vllm-cpp/Makefile

@@ -11,7 +11,7 @@ JOBS?=$(shell nproc --ignore=1 2>/dev/null || sysctl -n hw.ncpu 2>/dev/null || e
 
 # vllm.cpp version
 VLLM_CPP_REPO?=https://github.com/mudler/vllm.cpp
-VLLM_CPP_VERSION?=76f2a6d84e41cc9a593146a9a958f63645542731
+VLLM_CPP_VERSION?=6e99e4540b8d0811542a9d7690dc9aa0bda5812f
 
 # MLX GEMM provider (darwin/metal only; see the metal branch below for why).
 # Consumed as the prebuilt pip wheel: building MLX from source needs `xcrun

backend/go/whisper/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # whisper.cpp version
 WHISPER_REPO?=https://github.com/ggml-org/whisper.cpp
-WHISPER_CPP_VERSION?=1fe009caeda75f69bc864d6370b10674e45a92bd
+WHISPER_CPP_VERSION?=4834a2327d008ace3ec5a9ed00f51454bcabbc1c
 SO_TARGET?=libgowhisper.so
 
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF

backend/python/vllm/install.sh

@@ -119,7 +119,7 @@ if [ "$(uname -s)" = "Darwin" ]; then
     # can rewrite it. Darwin therefore follows vllm-metal and can lag the Linux
     # vllm pin (requirements-cublas13-after.txt, bumped independently against
     # vllm/vllm) until vllm-metal supports a newer vLLM.
-    VLLM_METAL_VERSION="v0.3.0.dev20260817081527"
+    VLLM_METAL_VERSION="v0.3.0.dev20260818075955"
 
     # The coupled vLLM source version is whatever this vllm-metal release builds
     # against. Derive it from

core/http/react-ui/e2e/stale-chunk-recovery.spec.js

@@ -12,7 +12,7 @@ import { test, expect } from './coverage-fixtures.js'
 // error instead of reloading forever.
 const HOME_CHUNK = /\/assets\/Home-[^/]*\.js(\?.*)?$/
 
-test.describe('Stale chunk recovery', () => {
+test.describe('@production-chunks Stale chunk recovery', () => {
   test('a 404 on a page chunk recovers on its own', async ({ page }) => {
     let served404 = 0
     await page.route(HOME_CHUNK, route => {

gallery/index.yaml

@@ -1,4 +1,55 @@
 ---
+- name: "qwen3.8-27b-heretic-abliterated-uncensored"
+  url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
+  urls:
+    - https://huggingface.co/0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored-GGUF
+  description: |
+    # Qwen3.8-27B
+
+    > [!Note]
+    > This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format.
+    >
+    > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, TokenSpeed, etc.
+
+    > [!Tip]
+    > For users seeking managed, scalable inference without infrastructure maintenance, the official Qwen API service is provided by Qwen Cloud.
+    > In particular, **Qwen3.8-27B** will be available as a hosted version with more production features, e.g., 1M context length by default, official built-in tools. For more information, please refer to the Qwen3.8-27B Overview. The service is coming soon. Stay tuned for updates.
+
+    Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date.
+
+    ...
+  license: "apache-2.0"
+  tags:
+    - llm
+    - gguf
+  icon: https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/CI_Demo/mathv-1327.jpg
+  overrides:
+    backend: llama-cpp
+    function:
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+    mmproj: llama-cpp/mmproj/Qwen3.8-27B-Heretic-Q4_K_M/mmproj-Qwen3.8-27B-Q8_0.gguf
+    options:
+      - use_jinja:true
+    parameters:
+      min_p: 0
+      model: llama-cpp/models/Qwen3.8-27B-Heretic-Q4_K_M/Qwen3.8-27B-Heretic-Q4_K_M.gguf
+      repeat_penalty: 1
+      temperature: 0.6
+      top_k: 20
+      top_p: 0.95
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/Qwen3.8-27B-Heretic-Q4_K_M/Qwen3.8-27B-Heretic-Q4_K_M.gguf
+      sha256: 5ae89d6720ba0a27f643e6fbb0d1b557870d17aeff23156e5981e863ec16985c
+      uri: https://huggingface.co/0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored-GGUF/resolve/main/Qwen3.8-27B-Heretic-Q4_K_M.gguf
+    - filename: llama-cpp/mmproj/Qwen3.8-27B-Heretic-Q4_K_M/mmproj-Qwen3.8-27B-Q8_0.gguf
+      sha256: 2e968a6af97ce35d8971890b257b9b7edabf20ad91450501fa53162a19ee33eb
+      uri: https://huggingface.co/0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored-GGUF/resolve/main/mmproj-Qwen3.8-27B-Q8_0.gguf
 - &ornith-1-0-9b
   name: "ornith-1.0-9b-q4"
   variants:
@@ -1591,6 +1642,74 @@
     - filename: llama-cpp/models/Parable-Qwen3-8B-Claude-Fable-5-Q4_K_M/Parable-Qwen3-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
       uri: https://huggingface.co/AnkitAI/Parable-Qwen3-8B-Claude-Fable-5-GGUF/resolve/main/Parable-Qwen3-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
       sha256: 4532d2379d38a37279866a030e51d419561f9d4d22fee00d2a33647d66f05065
+- &north-mini-code-1-0
+  name: "north-mini-code-1.0"
+  variants:
+    - model: north-mini-code-1.0-q8
+  url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
+  urls:
+    - https://huggingface.co/CohereLabs/North-Mini-Code-1.0
+    - https://huggingface.co/unsloth/North-Mini-Code-1.0-GGUF
+  description: |
+    North Mini Code 1.0 is Cohere Labs' Apache-2.0 sparse mixture-of-experts
+    coding model with 30B total parameters and 3B active parameters. It targets
+    code generation, agentic software engineering, terminal tasks, tool use,
+    and interleaved reasoning with a 256K-token context window. This entry uses
+    the UD-Q4_K_M GGUF quantization.
+  license: "apache-2.0"
+  tags:
+    - llm
+    - gguf
+    - cpu
+    - gpu
+    - moe
+    - coding
+    - agent
+  overrides:
+    backend: llama-cpp
+    function:
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-UD-Q4_K_M.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-UD-Q4_K_M.gguf
+      sha256: d59f4cb3abec1bd3e5e185b99cdd811e3be3402cbc2bab8223081a460d5724bf
+      uri: huggingface://unsloth/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-UD-Q4_K_M.gguf
+- !!merge <<: *north-mini-code-1-0
+  name: "north-mini-code-1.0-q8"
+  variants: []
+  description: |
+    North Mini Code 1.0 is Cohere Labs' Apache-2.0 sparse mixture-of-experts
+    coding model with 30B total parameters and 3B active parameters. It targets
+    code generation, agentic software engineering, terminal tasks, tool use,
+    and interleaved reasoning with a 256K-token context window. This entry uses
+    the Q8_0 GGUF quantization.
+  overrides:
+    backend: llama-cpp
+    function:
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-Q8_0.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-Q8_0.gguf
+      sha256: 652eeb9932e8020ef7b35374189411848187fa58796ded0ceee258bd9d36426b
+      uri: huggingface://unsloth/North-Mini-Code-1.0-GGUF/North-Mini-Code-1.0-Q8_0.gguf
 - &pocket-35b
   name: "pocket-35b"
   variants:
@@ -8102,6 +8221,94 @@
     - filename: llama-cpp/mmproj/BAAI_AREX-Turbo-GGUF/mmproj-BAAI_AREX-Turbo-f16.gguf
       sha256: 815184b9c47cac7d882b4b68aed98d610121c4ff38ce7ee0d015bdf3280f68a5
       uri: huggingface://bartowski/BAAI_AREX-Turbo-GGUF/mmproj-BAAI_AREX-Turbo-f16.gguf
+- &ui-mate-9b
+  name: ui-mate-9b
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  variants:
+    - model: ui-mate-9b-q8
+  urls:
+    - https://huggingface.co/tencent/UI-Mate-9B
+    - https://huggingface.co/bartowski/tencent_UI-Mate-9B-GGUF
+  description: |
+    UI-Mate-9B is Tencent's 9B-parameter multimodal computer-use agent,
+    fine-tuned from Qwen3.5-9B. It accepts task instructions, screenshots,
+    and interaction history, then emits reasoning and structured mouse and
+    keyboard actions for long-running desktop tasks. The model requires an
+    external runtime to execute its actions and should run with human
+    confirmation for sensitive operations. This entry uses the recommended
+    Q4_K_M GGUF quantization.
+  license: apache-2.0
+  tags:
+    - ui-mate
+    - qwen
+    - qwen3.5
+    - 9b
+    - llm
+    - gguf
+    - quantized
+    - chat
+    - vision
+    - multimodal
+    - agent
+    - computer-use
+    - gpu
+    - cpu
+  last_checked: "2026-08-18"
+  overrides:
+    backend: llama-cpp
+    function:
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+      - vision
+    mmproj: llama-cpp/mmproj/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q4_K_M.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q4_K_M.gguf
+      sha256: a54257eb4caa664dc1d2654b1aaed31767e3e6b01b4c7ecce9f239c00af23e2c
+      uri: huggingface://bartowski/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q4_K_M.gguf
+    - filename: llama-cpp/mmproj/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
+      sha256: 5a8380c4637dddceed9dbc28fffcdfa8601909c0ece9fe218fbd6888ec5d2c16
+      uri: huggingface://bartowski/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
+- !!merge <<: *ui-mate-9b
+  name: ui-mate-9b-q8
+  variants: []
+  description: |
+    UI-Mate-9B is Tencent's 9B-parameter multimodal computer-use agent,
+    fine-tuned from Qwen3.5-9B. It accepts task instructions, screenshots,
+    and interaction history, then emits reasoning and structured mouse and
+    keyboard actions for long-running desktop tasks. The model requires an
+    external runtime to execute its actions and should run with human
+    confirmation for sensitive operations. This entry uses the higher-quality
+    Q8_0 GGUF quantization.
+  overrides:
+    backend: llama-cpp
+    function:
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+      - vision
+    mmproj: llama-cpp/mmproj/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q8_0.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q8_0.gguf
+      sha256: 36f444d939188c904f4ce4449e9d3a690a081d26a828ada36b41df34ed5ad770
+      uri: huggingface://bartowski/tencent_UI-Mate-9B-GGUF/tencent_UI-Mate-9B-Q8_0.gguf
+    - filename: llama-cpp/mmproj/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
+      sha256: 5a8380c4637dddceed9dbc28fffcdfa8601909c0ece9fe218fbd6888ec5d2c16
+      uri: huggingface://bartowski/tencent_UI-Mate-9B-GGUF/mmproj-tencent_UI-Mate-9B-f16.gguf
 - &fara1-5-4b
   name: fara1.5-4b
   url: github:mudler/LocalAI/gallery/virtual.yaml@master