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

gallery/index.yaml

@@ -3295,7 +3295,7 @@
     - gemma3
     - gemma-3
   overrides:
     #mmproj: gemma-3-27b-it-mmproj-f16.gguf
     parameters:
       model: gemma-3-27b-it-Q4_K_M.gguf
   files:
@@ -3313,7 +3313,7 @@
   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:
@@ -3331,7 +3331,7 @@
   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:
@@ -6949,7 +6949,7 @@
       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
@@ -23023,3 +23023,41 @@
     - filename: Evilmind-24B-v1.i1-Q4_K_M.gguf
       sha256: 22e56c86b4f4a8f7eb3269f72a6bb0f06a7257ff733e21063fdec6691a52177d
       uri: huggingface://mradermacher/Evilmind-24B-v1-i1-GGUF/Evilmind-24B-v1.i1-Q4_K_M.gguf
+- !!merge <<: *llava
+  name: "gelato-30b-a3b-i1"
+  urls:
+    - https://huggingface.co/mradermacher/Gelato-30B-A3B-i1-GGUF
+  description: |
+    ### 🍨 Gelato-30B-A3B – A State-of-the-Art Vision-Language Model for GUI Grounding
+
+    **Overview**
+    Gelato-30B-A3B is a high-performance, open-source vision-language model (VLM) specifically designed for computer-use agent tasks. Trained on the large-scale **Click-100k** dataset, it excels at locating UI elements in graphical user interfaces (GUIs), making it ideal for automated interaction with software, web applications, and operating systems.
+
+    **Key Features**
+    - **Base Model**: Built upon **Qwen3-VL-30B-A3B-Instruct**, a powerful multimodal LLM with strong reasoning and vision capabilities.
+    - **Specialized Training**: Fine-tuned using data curation and reinforcement learning to achieve superior grounding accuracy.
+    - **High Accuracy**: Achieves **63.88% on ScreenSpot-Pro** and **73.40% on OS-World-G**, outperforming prior specialized models like GTA1-32B and even larger VLMs such as Qwen3-VL-235B.
+    - **Efficient Inference**: Activated size of only **3.3 GB**, enabling efficient deployment on consumer hardware.
+    - **Open Source & Free**: Fully open-access under the Apache 2.0 license with full training code and datasets available.
+
+    **Use Cases**
+    - Automating repetitive GUI interactions (e.g., form filling, software navigation)
+    - Building AI agents for desktop and web automation
+    - Research in computer-use agent behavior and human-AI collaboration
+
+    **Inference Example**
+    Given a screen image and a natural language instruction like *"Reload the cache"*, Gelato outputs precise (x,y) coordinates of the target UI element—enabling accurate mouse clicks or touch actions.
+
+    **Model Link**
+    👉 [View on Hugging Face: mlfoundations-cua-dev/Gelato-30B-A3B](https://huggingface.co/mlfoundations-cua-dev/Gelato-30B-A3B)
+
+    **Ideal For**
+    Developers, AI researchers, and automation engineers seeking a lightweight, high-accuracy model for GUI interaction and agent-based tasks.
+    *Bonus*: When paired with GPT-5, it enables frontier-level agentic performance on OS-World.
+  overrides:
+    parameters:
+      model: Gelato-30B-A3B.i1-Q4_K_M.gguf
+  files:
+    - filename: Gelato-30B-A3B.i1-Q4_K_M.gguf
+      sha256: b353b25d0e193340dbf68261d930f5456adb2933a85d74be5296757d85337f45
+      uri: huggingface://mradermacher/Gelato-30B-A3B-i1-GGUF/Gelato-30B-A3B.i1-Q4_K_M.gguf