chore(gallery): add Mellum2.1 Thinking variants

Add JetBrains Q4_K_M and Q8_0 GGUF builds for llama.cpp, with
variant selection and installation documentation. Pin both downloads
and verify their SHA256 values against LFS metadata and HTTP headers.

Assisted-by: Codex:gpt-6

docs/content/features/model-gallery.md

@@ -902,3 +902,27 @@ is still running is reported as queued until the installer picks it up:
 A job ID is queryable from the moment `/models/apply` returns it, so a `404`/`500`
 from this endpoint means the ID is genuinely unknown rather than merely waiting
 its turn.
+
+### Mellum2.1 Thinking
+
+[Mellum2.1 Thinking](https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking)
+is a JetBrains reasoning model with 12 billion total parameters and 2.5 billion
+active parameters per token. The gallery provides Q4_K_M and Q8_0 GGUF builds
+for the llama-cpp backend.
+
+Install the default entry with automatic variant selection:
+
+```bash
+local-ai models install mellum2.1-12b-a2.5b-thinking
+```
+
+To select Q8_0 explicitly:
+
+```bash
+local-ai models install mellum2.1-12b-a2.5b-thinking --variant mellum2.1-12b-a2.5b-thinking-q8
+```
+
+Both entries use the embedded chat template, temperature 0.6, top-p 0.95,
+and top-k 20. The default context is 32,768 tokens; the model supports up to
+131,072 tokens. Increasing the context requires more memory. The model emits
+reasoning inside `<think>...</think>` blocks before its final answer.

gallery/index.yaml

@@ -8766,6 +8766,82 @@
     - filename: llama-cpp/models/FINAL-Bench/POCKET-26B-GGUF/POCKET-26B-Q2_K.gguf
       sha256: 490687be6493769dc8322df396cd31d412a15877eaccea4d2f16ca1f93383334
       uri: huggingface://FINAL-Bench/POCKET-26B-GGUF/POCKET-26B-Q2_K.gguf
+- name: mellum2.1-12b-a2.5b-thinking
+  variants:
+    - model: mellum2.1-12b-a2.5b-thinking-q8
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking
+    - https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF
+  description: |
+    Mellum2.1 Thinking is a JetBrains mixture-of-experts reasoning model with
+    12 billion total parameters and 2.5 billion active parameters per token.
+    It emits reasoning in think blocks and supports a 131,072-token context.
+    This entry uses Q4_K_M GGUF weights with a 32,768-token default context.
+  license: apache-2.0
+  tags:
+    - llm
+    - gguf
+    - cpu
+    - gpu
+    - moe
+    - reasoning
+  last_checked: "2026-10-08"
+  overrides:
+    backend: llama-cpp
+    context_size: 32768
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/JetBrains/Mellum2.1-12B-A2.5B-Thinking/Mellum2.1-12B-A2.5B-Thinking-Q4_K_M.gguf
+      temperature: 0.6
+      top_p: 0.95
+      top_k: 20
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/JetBrains/Mellum2.1-12B-A2.5B-Thinking/Mellum2.1-12B-A2.5B-Thinking-Q4_K_M.gguf
+      sha256: ecc4d5b8107fc219e23c494b2d107552137637b6c6002ee713e0ba8af7fa2755
+      uri: https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF/resolve/20b439426f1e997ba575eaccaf4856ac0f76f5d5/Mellum2.1-12B-A2.5B-Thinking-Q4_K_M.gguf
+- name: mellum2.1-12b-a2.5b-thinking-q8
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking
+    - https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF
+  description: |
+    Mellum2.1 Thinking is a JetBrains mixture-of-experts reasoning model with
+    12 billion total parameters and 2.5 billion active parameters per token.
+    It emits reasoning in think blocks and supports a 131,072-token context.
+    This entry uses Q8_0 GGUF weights with a 32,768-token default context.
+  license: apache-2.0
+  tags:
+    - llm
+    - gguf
+    - cpu
+    - gpu
+    - moe
+    - reasoning
+  last_checked: "2026-10-08"
+  overrides:
+    backend: llama-cpp
+    context_size: 32768
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/JetBrains/Mellum2.1-12B-A2.5B-Thinking/Mellum2.1-12B-A2.5B-Thinking-Q8_0.gguf
+      temperature: 0.6
+      top_p: 0.95
+      top_k: 20
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/JetBrains/Mellum2.1-12B-A2.5B-Thinking/Mellum2.1-12B-A2.5B-Thinking-Q8_0.gguf
+      sha256: 69e0ac02a16586c2c9e991f1e3b8a1eb403882e7ab9e69ea01eb23dbbbc6841f
+      uri: https://huggingface.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF/resolve/20b439426f1e997ba575eaccaf4856ac0f76f5d5/Mellum2.1-12B-A2.5B-Thinking-Q8_0.gguf
 - &mellum2-12b-a2-5b-instruct
   name: mellum2-12b-a2.5b-instruct
   variants: