chore(gallery): add B1-9B GGUF variants

Add Q4_K_M and Q8_0 builds with the vision projector for llama.cpp.
Pin artifacts to a verified revision and document installation.

Assisted-by: Codex:gpt-6
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

docs/content/features/model-gallery.md

@@ -286,6 +286,27 @@ These entries use `stablediffusion-ggml` and include the text encoder, vision pr
 The invalid `qwen-image-2.1-uncensored` chat entry was removed because llama.cpp cannot load its diffusion weights.
 This removal does not delete previously installed models. Remove that configuration before installing an image-generation entry.
 
+## B1-9B
+
+B1-9B is a Qwen3.5-based model from Schneewolf Labs for reasoning and tool use.
+The gallery provides Q4_K_M and Q8_0 builds for llama.cpp, both with a Q8_0
+vision projector and an 8192-token default context.
+
+Install the default build, or select Q8_0 explicitly:
+
+```bash
+local-ai models install b1-9b
+local-ai models install --variant b1-9b-q8 b1-9b
+```
+
+Send tool definitions through the API's `tools` field. The publisher recommends
+avoiding a `/think` prompt prefix for tool-result turns. These entries use the
+embedded chat template and do not enable speculative decoding.
+
+See the [model card](https://huggingface.co/schneewolflabs/B1-9B) for training details
+and limitations, and the [GGUF repository](https://huggingface.co/mradermacher/B1-9B-GGUF)
+for quantization files.
+
 ## VRAM and download size estimates
 
 When browsing the gallery or importing a model by URI, LocalAI can show **estimated download size** and **estimated VRAM** for models.

gallery/index.yaml

@@ -7349,6 +7349,84 @@
     - filename: llama-cpp/models/homura-30b/Homura-30B-Q4_K_M.gguf
       uri: huggingface://hyrelabs/Homura-30B-GGUF/Homura-30B-Q4_K_M.gguf
       sha256: fd4dc394b193ac1cc0f4bb9fb44bff3d6eebd34df62246fc753104c0cc98cc51
+- &b1-9b
+  name: b1-9b
+  variants:
+    - model: b1-9b-q8
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/schneewolflabs/B1-9B
+    - https://huggingface.co/mradermacher/B1-9B-GGUF
+  description: |
+    B1-9B is Schneewolf Labs' Qwen3.5-based 9B model for reasoning and tool use.
+    It is trained to produce an answer after reasoning about tool results.
+    This Q4_K_M GGUF build includes the Q8_0 vision projector and uses an
+    8192-token context by default. A Q8_0 variant is also available.
+  license: apache-2.0
+  tags:
+    - llm
+    - gguf
+    - cpu
+    - gpu
+    - qwen
+    - reasoning
+    - tools
+    - vision
+  last_checked: "2026-10-11"
+  overrides:
+    backend: llama-cpp
+    context_size: 8192
+    mmproj: llama-cpp/mmproj/b1-9b/B1-9B.mmproj-Q8_0.gguf
+    function:
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+      - vision
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/b1-9b/B1-9B.Q4_K_M.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/b1-9b/B1-9B.Q4_K_M.gguf
+      uri: https://huggingface.co/mradermacher/B1-9B-GGUF/resolve/80720551357afc6f6933e4de206f9e58e06653a1/B1-9B.Q4_K_M.gguf
+      sha256: 61576c9c0d471d5d4ce5d8a04cad1efac74669eeaa059c49a96a55bc38143786
+    - filename: llama-cpp/mmproj/b1-9b/B1-9B.mmproj-Q8_0.gguf
+      uri: https://huggingface.co/mradermacher/B1-9B-GGUF/resolve/80720551357afc6f6933e4de206f9e58e06653a1/B1-9B.mmproj-Q8_0.gguf
+      sha256: ef20f8a1ce884ddee5155b79d8785bd0400bd0350d5a31dd330cb21c69b09931
+- !!merge <<: *b1-9b
+  name: b1-9b-q8
+  variants: []
+  description: |
+    B1-9B in Q8_0 GGUF format with the Q8_0 vision projector. This variant
+    preserves more weight precision and requires more memory than Q4_K_M.
+  overrides:
+    backend: llama-cpp
+    context_size: 8192
+    mmproj: llama-cpp/mmproj/b1-9b/B1-9B.mmproj-Q8_0.gguf
+    function:
+      automatic_tool_parsing_fallback: true
+      grammar:
+        disable: true
+    known_usecases:
+      - chat
+      - vision
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/b1-9b/B1-9B.Q8_0.gguf
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/b1-9b/B1-9B.Q8_0.gguf
+      uri: https://huggingface.co/mradermacher/B1-9B-GGUF/resolve/80720551357afc6f6933e4de206f9e58e06653a1/B1-9B.Q8_0.gguf
+      sha256: d43e20632c62748df9d67aaeb336168f383a78021f6e74b2bb9dcf9925116d3f
+    - filename: llama-cpp/mmproj/b1-9b/B1-9B.mmproj-Q8_0.gguf
+      uri: https://huggingface.co/mradermacher/B1-9B-GGUF/resolve/80720551357afc6f6933e4de206f9e58e06653a1/B1-9B.mmproj-Q8_0.gguf
+      sha256: ef20f8a1ce884ddee5155b79d8785bd0400bd0350d5a31dd330cb21c69b09931
 - &qwen3-5-9b-defiant-fable
   name: qwen3.5-9b-defiant-fable-mtp
   variants: