fix(gallery): use native Ministral tool parsing (#12022)

The Ministral 3 14B Reasoning entry inherits a Mistral 0.3 prompt and
JSON parser. Its name-first tool calls can therefore reach clients as
plain text.

Use the embedded template and llama.cpp's native tool parser. Document
migration for installed configurations, which gallery updates do not
rewrite.

Assisted-by: Codex:gpt-6

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Co-authored-by: localai-org-maint-bot <306269227+localai-org-maint-bot@users.noreply.github.com>

docs/content/features/openai-functions.md

@@ -36,6 +36,34 @@ Reasoning content (`<think>...</think>` blocks from DeepSeek R1, Qwen3, Gemma 4,
 
 No configuration required - the autoparser detects the tool call format for any `ggml`/`gguf` model that was trained with tool support.
 
+#### Migrating older Ministral 3 configurations
+
+The `mistralai_ministral-3-14b-reasoning-2512-multimodal` gallery entry uses the model's embedded template and llama.cpp's native tool parser.
+Older installations inherit the Mistral 0.3 prompt and JSON parser, which can return tool calls such as `pick_tool{...}` as text.
+
+For an existing installation, replace the model YAML's `template` and `function` sections with:
+
+```yaml
+template:
+  use_tokenizer_template: true
+function:
+  disable_no_action: true
+  automatic_tool_parsing_fallback: true
+  grammar:
+    disable: true
+```
+
+Set `use_jinja:true` in the existing `options` list:
+
+```yaml
+options:
+  - use_jinja:true
+```
+
+Remove the inherited `stopwords` list so llama.cpp controls the model's end-of-turn markers.
+Keep your model path, `mmproj`, sampling settings, and MCP configuration. Reload the model after saving the YAML.
+Gallery changes do not rewrite installed model configurations.
+
 ### vLLM / vLLM Omni
 
 The parser must be specified explicitly because vLLM itself doesn't auto-detect one. Pass it via the model `options`:
@@ -312,4 +340,4 @@ Grammars and function tools can be used as well in conjunction with vision APIs:
 
 ## 💡 Examples
 
-A full e2e example with `docker-compose` is available [here](https://github.com/mudler/LocalAI-examples/tree/main/functions).
+A full e2e example with `docker-compose` is available [here](https://github.com/mudler/LocalAI-examples/tree/main/functions).

gallery/index.yaml

@@ -41311,7 +41311,7 @@
       sha256: 2740ba9e9b30b09be4282a9a9f617ec43dc47b89aed416cb09b5f698f90783b5
       uri: huggingface://unsloth/Ministral-3-14B-Instruct-2512-GGUF/mmproj-F32.gguf
 - name: mistralai_ministral-3-14b-reasoning-2512-multimodal
-  url: github:mudler/LocalAI/gallery/mistral-0.3.yaml@master
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
   urls:
     - https://huggingface.co/mistralai/Ministral-3-14B-Reasoning-2512
     - https://huggingface.co/unsloth/Ministral-3-14B-Reasoning-2512-GGUF
@@ -41359,8 +41359,19 @@
     - vision
   last_checked: "2026-05-04"
   overrides:
+    backend: llama-cpp
     context_size: 32768
+    mmap: true
     mmproj: llama-cpp/mmproj/mmproj-mistralai_Ministral-3-14B-Reasoning-2512-f32.gguf
+    options:
+      - use_jinja:true
+    template:
+      use_tokenizer_template: true
+    function:
+      disable_no_action: true
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
     parameters:
       model: llama-cpp/models/mistralai_Ministral-3-14B-Reasoning-2512-Q4_K_M.gguf
       temperature: 0.7