chore(model gallery): 🤖 add 1 new models via gallery agent

gallery/index.yaml

@@ -3127,7 +3127,7 @@
     - gemma3
     - gemma-3
   overrides:
     #mmproj: gemma-3-27b-it-mmproj-f16.gguf
     parameters:
       model: gemma-3-27b-it-Q4_K_M.gguf
   files:
@@ -3145,7 +3145,7 @@
   description: |
     google/gemma-3-12b-it is an open-source, state-of-the-art, lightweight, multimodal model built from the same research and technology used to create the Gemini models. It is capable of handling text and image input and generating text output. It has a large context window of 128K tokens and supports over 140 languages. The 12B variant has been fine-tuned using the instruction-tuning approach. Gemma 3 models are suitable for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning. Their relatively small size makes them deployable in environments with limited resources such as laptops, desktops, or your own cloud infrastructure.
   overrides:
     #mmproj: gemma-3-12b-it-mmproj-f16.gguf
     parameters:
       model: gemma-3-12b-it-Q4_K_M.gguf
   files:
@@ -3163,7 +3163,7 @@
   description: |
     Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. Gemma 3 models are multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned variants. Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions. Gemma 3 models are well-suited for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning. Their relatively small size makes it possible to deploy them in environments with limited resources such as laptops, desktops or your own cloud infrastructure, democratizing access to state of the art AI models and helping foster innovation for everyone. Gemma-3-4b-it is a 4 billion parameter model.
   overrides:
     #mmproj: gemma-3-4b-it-mmproj-f16.gguf
     parameters:
       model: gemma-3-4b-it-Q4_K_M.gguf
   files:
@@ -6781,7 +6781,7 @@
       sha256: 2756551de7d8ff7093c2c5eec1cd00f1868bc128433af53f5a8d434091d4eb5a
       uri: huggingface://Triangle104/Nano_Imp_1B-Q8_0-GGUF/nano_imp_1b-q8_0.gguf
 - &qwen25
   name: "qwen2.5-14b-instruct" ## Qwen2.5
   icon: https://avatars.githubusercontent.com/u/141221163
   url: "github:mudler/LocalAI/gallery/chatml.yaml@master"
   license: apache-2.0
@@ -23150,3 +23150,41 @@
     - filename: financial-gpt-oss-20b-q8.i1-Q4_K_M.gguf
       sha256: 14586673de2a769f88bd51f88464b9b1f73d3ad986fa878b2e0c1473f1c1fc59
       uri: huggingface://mradermacher/financial-gpt-oss-20b-q8-i1-GGUF/financial-gpt-oss-20b-q8.i1-Q4_K_M.gguf
+- !!merge <<: *llama3
+  name: "qwen3-vl-235b-a22b-instruct"
+  urls:
+    - https://huggingface.co/unsloth/Qwen3-VL-235B-A22B-Instruct-GGUF
+  description: |
+    **Qwen3-VL-235B-A22B-Instruct**
+    *by Qwen Team (Alibaba Cloud)*
+
+    A state-of-the-art vision-language model in the Qwen series, Qwen3-VL is the most powerful multimodal model to date. It combines advanced visual perception with exceptional language understanding, enabling deep reasoning across images, videos, and text.
+
+    **Key Features:**
+    - **235B parameters** with MoE (Mixture of Experts) architecture for scalable performance.
+    - Supports **256K native context length** (expandable to 1M), ideal for long documents and video analysis.
+    - **Advanced multimodal reasoning** in STEM, logic, and visual coding (e.g., generates HTML/CSS/JS from images).
+    - **Enhanced spatial and temporal understanding** with precise object localization and timestamp-aware video analysis.
+    - **Visual agent capabilities** for interacting with UIs and performing tasks on PC/mobile devices.
+    - **32-language OCR** with high accuracy under challenging conditions (low light, blur, tilt).
+    - Seamlessly integrates text and vision for unified, lossless comprehension.
+
+    **Architecture Innovations:**
+    - **Interleaved-MRoPE** for superior long-horizon video reasoning.
+    - **DeepStack** for fine-grained visual-text alignment.
+    - **Text–Timestamp Alignment** for precise event localization in videos.
+
+    **Best For:** Complex multimodal tasks including visual reasoning, document analysis, video understanding, and agent-based interactions.
+
+    **Model ID:** `Qwen/Qwen3-VL-235B-A22B-Instruct`
+    **License:** Apache 2.0
+    **Official Repository:** [https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct)
+
+    > 📌 Note: This is the *original*, non-quantized model from the official Qwen team. The GGUF version (e.g., `unsloth/Qwen3-VL-235B-A22B-Instruct-GGUF`) is a quantized variant for efficient inference and should not be confused with the base model.
+  overrides:
+    parameters:
+      model: UD-Q2_K_XL
+  files:
+    - filename: UD-Q2_K_XL
+      sha256: 3249c3bf674a2e8bc942e04a2c8cc9d9f9e4a4c8
+      uri: huggingface://unsloth/Qwen3-VL-235B-A22B-Instruct-GGUF/UD-Q2_K_XL