@tecet/ollm 0.1.3

package/dist/config/LLM_profiles.json

@@ -0,0 +1,1908 @@
+{
+    "version": "0.2.0",
+    "models": [
+        {
+            "id": "user-unknown-model",
+            "name": "Unknown Model",
+            "creator": "User",
+            "parameters": "Based on Llama 3.2 3B",
+            "quantization": "Based on Llama 3.2 3B (4-bit estimated)",
+            "description": "Unknown model - please edit your settings at ~/.ollm/LLM_profiles.json (Windows: C:\\Users\\{username}\\.ollm\\LLM_profiles.json)",
+            "abilities": ["Unknown"],
+            "tool_support": false,
+            "ollama_url": "https://github.com/Tecet/OLLM/tree/main/docs",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "2.5 GB",
+                    "ollama_context_size": 2867,
+                    "vram_estimate_gb": 2.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "2.9 GB",
+                    "ollama_context_size": 5734,
+                    "vram_estimate_gb": 2.9
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "3.7 GB",
+                    "ollama_context_size": 11468,
+                    "vram_estimate_gb": 3.7
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "5.2 GB",
+                    "ollama_context_size": 22937,
+                    "vram_estimate_gb": 5.2
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "8.2 GB",
+                    "ollama_context_size": 45875,
+                    "vram_estimate_gb": 8.2
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "14.2 GB",
+                    "ollama_context_size": 91750,
+                    "vram_estimate_gb": 14.2
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "codegeex4:9b",
+            "name": "CodeGeeX4 9B",
+            "creator": "Zhipu AI",
+            "parameters": "9B",
+            "quantization": "Q4_0",
+            "size": "5.5 GB",
+            "description": "Multilingual code generation model with high inference speed.",
+            "abilities": ["Coding", "Multilingual"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/codegeex4",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.1 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.1
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "9.2 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 9.2
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "11.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 11.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.8 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.8
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "codegemma:2b",
+            "name": "CodeGemma 2B",
+            "creator": "Google DeepMind",
+            "parameters": "2B",
+            "quantization": "Q4_K_M",
+            "size": "1.6 GB",
+            "description": "Lightweight code completion model for on-device use.",
+            "abilities": ["Coding", "Code Completion"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/codegemma",
+            "max_context_window": 8192,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "3.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 3.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "3.4 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 3.4
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "codegemma:7b",
+            "name": "CodeGemma 7B",
+            "creator": "Google DeepMind",
+            "parameters": "7B",
+            "quantization": "Q4_K_M",
+            "size": "5.0 GB",
+            "description": "Specialized code generation and completion model.",
+            "abilities": ["Coding", "Code Generation"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/codegemma",
+            "max_context_window": 8192,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "codestral:22b",
+            "name": "Codestral 22B",
+            "creator": "Mistral AI",
+            "parameters": "22B",
+            "quantization": "Q4_0",
+            "size": "12 GB",
+            "description": "Mistral's model optimized for intermediate code generation tasks.",
+            "abilities": ["Coding", "Code Generation"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/codestral",
+            "max_context_window": 32768,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "14.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 14.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "15.6 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 15.6
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "17.8 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 17.8
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "22.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 22.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "command-r7b:7b",
+            "name": "Command R7B 7B",
+            "creator": "Cohere",
+            "parameters": "7B",
+            "quantization": "Q4_K_M",
+            "size": "5.1 GB",
+            "description": "Optimized for RAG and tool use, excellent instruction following.",
+            "abilities": ["General Purpose", "RAG", "Tool Use"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/command-r7b",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.6 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.6
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.8 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.8
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "11.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 11.0
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "deepseek-coder-v2:16b",
+            "name": "DeepSeek Coder V2 16B",
+            "creator": "DeepSeek",
+            "parameters": "16B",
+            "quantization": "Q4_K_M",
+            "size": "8.9 GB",
+            "description": "Mixture-of-Experts (MoE) model for advanced coding tasks.",
+            "abilities": ["Coding", "MoE Architecture"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/deepseek-coder-v2",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "11.2 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 11.2
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "11.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 11.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "12.0 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 12.0
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "13.1 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 13.1
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.2 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.2
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "19.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 19.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "deepseek-r1:1.5b",
+            "name": "DeepSeek R1 1.5B",
+            "creator": "DeepSeek",
+            "parameters": "1.5B",
+            "quantization": "Q4_K_M",
+            "size": "1.1 GB",
+            "description": "Distilled reasoning model, highly efficient for logic puzzles.",
+            "abilities": ["Reasoning", "Math", "Logic"],
+            "tool_support": false,
+            "thinking_enabled": true,
+            "reasoning_buffer": "Variable",
+            "warmup_timeout": 120000,
+            "ollama_url": "https://ollama.com/library/deepseek-r1",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "2.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 2.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "2.8 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 2.8
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "3.2 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 3.2
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "4.1 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 4.1
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "6.0 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 6.0
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "9.8 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 9.8
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "deepseek-r1:7b",
+            "name": "DeepSeek R1 7B",
+            "creator": "DeepSeek",
+            "parameters": "7B",
+            "quantization": "Q4_K_M",
+            "size": "4.7 GB",
+            "description": "Distilled reasoning model based on Qwen 2.5, excels in math/logic.",
+            "abilities": ["Reasoning", "Math", "Logic"],
+            "tool_support": false,
+            "thinking_enabled": true,
+            "reasoning_buffer": "Variable",
+            "warmup_timeout": 120000,
+            "ollama_url": "https://ollama.com/library/deepseek-r1",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.2
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.4 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.4
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "10.8 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 10.8
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "deepseek-r1:8b",
+            "name": "DeepSeek R1 8B",
+            "creator": "DeepSeek",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "5.2 GB",
+            "description": "Distilled reasoning model based on Llama 3, balanced for logic.",
+            "abilities": ["Reasoning", "Math", "Logic"],
+            "tool_support": false,
+            "thinking_enabled": true,
+            "reasoning_buffer": "Variable",
+            "warmup_timeout": 120000,
+            "ollama_url": "https://ollama.com/library/deepseek-r1",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.8 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.8
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "9.0 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 9.0
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "11.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 11.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "16.2 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 16.2
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "25.0 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 25.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "dolphin3:8b",
+            "name": "Dolphin3 8B",
+            "creator": "Cognitive Computations",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "4.9 GB",
+            "description": "Uncensored/compliant model optimized for general conversation/coding.",
+            "abilities": ["General Purpose", "Coding", "Uncensored"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/dolphin3",
+            "max_context_window": 32768,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.7 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.7
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.9 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.9
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "11.2 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 11.2
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "gemma3:1b",
+            "name": "Gemma 3 1B",
+            "creator": "Google DeepMind",
+            "parameters": "1B",
+            "quantization": "Q4_K_M",
+            "size": "815 MB",
+            "description": "Ultra-compact text model from Google, high efficiency.",
+            "abilities": ["General Purpose", "Fast Inference"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/gemma3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "2.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 2.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "2.4 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 2.4
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "3.2 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 3.2
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "4.8 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 4.8
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "8.0 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 8.0
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "14.2 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 14.2
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "gemma3:4b",
+            "name": "Gemma 3 4B",
+            "creator": "Google DeepMind",
+            "parameters": "4B",
+            "quantization": "Q4_K_M",
+            "size": "3.3 GB",
+            "description": "Balanced Google model with strong general purpose capabilities.",
+            "abilities": ["General Purpose", "Reasoning"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/gemma3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "5.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 5.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "5.8 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 5.8
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "10.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 10.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "16.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 16.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "28.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 28.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "gemma3n:e2b",
+            "name": "Gemma 3 Nano E2B",
+            "creator": "Google DeepMind",
+            "parameters": "E2B",
+            "quantization": "Q4_K_M",
+            "size": "5.6 GB",
+            "description": "Nano/Experimental variant of Gemma 3.",
+            "abilities": ["General Purpose", "Experimental"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/gemma3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "10.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 10.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "14.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 14.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "22.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 22.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "38.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 38.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "gemma3n:latest",
+            "name": "Gemma 3 Nano Latest",
+            "creator": "Google DeepMind",
+            "parameters": "Latest",
+            "quantization": "Q4_K_M",
+            "size": "7.5 GB",
+            "description": "Latest standard variant of the Gemma 3 Nano/Next series.",
+            "abilities": ["General Purpose"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/gemma3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "9.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 9.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "10.8 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 10.8
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "13.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 13.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "18.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 18.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "28.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 28.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "48.0 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 48.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "gpt-oss:20b",
+            "name": "GPT-OSS 20B",
+            "creator": "OpenAI (Hypothetical/Mirror)",
+            "parameters": "20B",
+            "quantization": "Q4_0",
+            "size": "13 GB",
+            "description": "Open-weight model by OpenAI (hypothetical/leaked) for general tasks.",
+            "abilities": ["General Purpose"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/gpt-oss",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "16.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 16.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "17.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 17.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "19.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 19.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "23.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 23.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "31.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 31.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "47.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 47.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "granite-code:3b",
+            "name": "Granite Code 3B",
+            "creator": "IBM",
+            "parameters": "3B",
+            "quantization": "Q4_K_M",
+            "size": "2.0 GB",
+            "description": "IBM enterprise coding model, very fast.",
+            "abilities": ["Coding"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/granite-code",
+            "max_context_window": 4096,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "3.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 3.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "granite-code:8b",
+            "name": "Granite Code 8B",
+            "creator": "IBM",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "4.6 GB",
+            "description": "IBM enterprise coding model, robust for Python/Java/JS.",
+            "abilities": ["Coding"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/granite-code",
+            "max_context_window": 4096,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "granite3.3:8b",
+            "name": "Granite 3.3 8B",
+            "creator": "IBM",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "4.9 GB",
+            "description": "Updated IBM model with improved reasoning and instruction following.",
+            "abilities": ["General Purpose", "Reasoning"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/granite3.3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.7 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.7
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.9 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.9
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "11.2 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 11.2
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.8 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.8
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.8 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.8
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "granite4:latest",
+            "name": "Granite 4 Latest",
+            "creator": "IBM",
+            "parameters": "Latest",
+            "quantization": "Q4_K_M",
+            "size": "2.1 GB",
+            "description": "Next-gen IBM model, hybrid Mamba architecture for speed.",
+            "abilities": ["General Purpose", "Fast Inference"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/granite4",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "4.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 4.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "4.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 4.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "5.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 5.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "11.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 11.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "19.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 19.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "llama3:latest",
+            "name": "Llama 3 Latest",
+            "creator": "Meta",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "4.7 GB",
+            "description": "Meta's 8B standard model (robust generalist).",
+            "abilities": ["General Purpose"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/llama3",
+            "max_context_window": 8192,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.2
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "llama3.2:3b",
+            "name": "Llama 3.2 3B",
+            "creator": "Meta",
+            "parameters": "3B",
+            "quantization": "Q4_K_M",
+            "size": "2.0 GB",
+            "description": "Lightweight Llama optimized for edge devices.",
+            "abilities": ["General Purpose", "Edge Optimized"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/llama3.2",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "3.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 3.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "3.9 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 3.9
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "4.6 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 4.6
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "6.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 6.0
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "8.8 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 8.8
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "14.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 14.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "magicoder:latest",
+            "name": "Magicoder Latest",
+            "creator": "ise-uiuc (OSS-Instruct Team)",
+            "parameters": "Latest",
+            "quantization": "Q4_K_M",
+            "size": "3.8 GB",
+            "description": "Code model trained on synthetic data (OSS-Instruct).",
+            "abilities": ["Coding"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/magicoder",
+            "max_context_window": 16384,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "5.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 5.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "6.0 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 6.0
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "7.1 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 7.1
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "ministral-3:3b",
+            "name": "Ministral 3 3B",
+            "creator": "Mistral AI",
+            "parameters": "3B",
+            "quantization": "Q4_K_M",
+            "size": "3.0 GB",
+            "description": "Mistral's edge model with high intelligence-to-size ratio.",
+            "abilities": ["General Purpose", "Edge Optimized"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/ministral-3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "5.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 5.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "5.6 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 5.6
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "6.8 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 6.8
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "9.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 9.0
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "13.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 13.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "22.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 22.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "ministral-3:8b",
+            "name": "Ministral 3 8B",
+            "creator": "Mistral AI",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "6.0 GB",
+            "description": "Larger edge model from Mistral, strong instruction following.",
+            "abilities": ["General Purpose", "Instruction Following"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/ministral-3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "8.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 8.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.8 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.8
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "10.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 10.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "13.6 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 13.6
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "20.0 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 20.0
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "32.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 32.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "opencoder:8b",
+            "name": "OpenCoder 8B",
+            "creator": "OpenCoder Team / INF",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "4.7 GB",
+            "description": "Fully open-source coding model with transparent dataset.",
+            "abilities": ["Coding"],
+            "tool_support": false,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/opencoder",
+            "max_context_window": 8192,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.2
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "phi3:3.8b",
+            "name": "Phi-3 Mini 3.8B",
+            "creator": "Microsoft",
+            "parameters": "3.8B",
+            "quantization": "Q4_K_M",
+            "size": "2.2 GB",
+            "description": "Microsoft's highly capable small model (Mini).",
+            "abilities": ["Reasoning", "Coding"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/phi3",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "4.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 4.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "4.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 4.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "5.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 5.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "11.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 11.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "19.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 19.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "phi4-mini-reasoning:latest",
+            "name": "Phi-4 Mini Reasoning",
+            "creator": "Microsoft",
+            "parameters": "Latest",
+            "quantization": "Q4_K_M",
+            "size": "3.2 GB",
+            "description": "Microsoft model specialized in math and logical reasoning steps.",
+            "abilities": ["Reasoning", "Math", "Logic"],
+            "tool_support": false,
+            "thinking_enabled": true,
+            "ollama_url": "https://ollama.com/library/phi4",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "5.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 5.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "5.6 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 5.6
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "6.8 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 6.8
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "9.2 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 9.2
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "13.8 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 13.8
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "23.0 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 23.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen2.5:7b",
+            "name": "Qwen 2.5 7B",
+            "creator": "Alibaba Cloud",
+            "parameters": "7B",
+            "quantization": "Q4_K_M",
+            "size": "4.7 GB",
+            "description": "Alibaba's strong generalist, beats Llama 3.1 8B in benchmarks.",
+            "abilities": ["General Purpose", "Coding", "Math", "Multilingual"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen2.5",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.2
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.4 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.4
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "10.8 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 10.8
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen2.5-coder:3b",
+            "name": "Qwen 2.5 Coder 3B",
+            "creator": "Alibaba Cloud",
+            "parameters": "3B",
+            "quantization": "Q4_K_M",
+            "size": "1.9 GB",
+            "description": "Small but potent coding assistant.",
+            "abilities": ["Coding"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen2.5-coder",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "3.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 3.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "4.0 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 4.0
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "4.8 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 4.8
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "9.8 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 9.8
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "16.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 16.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen2.5-coder:7b",
+            "name": "Qwen 2.5 Coder 7B",
+            "creator": "Alibaba Cloud",
+            "parameters": "7B",
+            "quantization": "Q4_K_M",
+            "size": "4.7 GB",
+            "description": "SOTA coding model in the 7B class.",
+            "abilities": ["Coding"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen2.5-coder",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "7.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 7.2
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "8.4 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 8.4
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "10.8 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 10.8
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "15.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 15.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 24.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen3:4b",
+            "name": "Qwen 3 4B",
+            "creator": "Alibaba Cloud",
+            "parameters": "4B",
+            "quantization": "Q4_K_M",
+            "size": "2.5 GB",
+            "description": "Next-gen Qwen generalist, high efficiency.",
+            "abilities": ["General Purpose", "Coding", "Reasoning"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen3",
+            "max_context_window": 262144,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "4.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 4.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "5.0 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 5.0
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "6.0 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 6.0
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "8.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 8.0
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "12.0 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 12.0
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "20.0 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 20.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen3:8b",
+            "name": "Qwen 3 8B",
+            "creator": "Alibaba Cloud",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "5.2 GB",
+            "description": "Next-gen Qwen generalist 8B, high reasoning capability.",
+            "abilities": ["General Purpose", "Coding", "Reasoning"],
+            "tool_support": true,
+            "thinking_enabled": true,
+            "ollama_url": "https://ollama.com/library/qwen3",
+            "max_context_window": 262144,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "7.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 7.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.2 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.2
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "9.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 9.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "12.0 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 12.0
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "17.0 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 17.0
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "27.0 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 27.0
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen3-coder:30b",
+            "name": "Qwen 3 Coder 30B",
+            "creator": "Alibaba Cloud",
+            "parameters": "30B",
+            "quantization": "Q4_0",
+            "size": "18 GB",
+            "description": "Heavyweight coding model, expert level generation.",
+            "abilities": ["Coding"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen3-coder",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "20.5 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 20.5
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "22.0 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 22.0
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "24.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 24.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "29.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 29.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "39.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 39.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "59.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 59.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen3-vl:4b",
+            "name": "Qwen 3 VL 4B",
+            "creator": "Alibaba Cloud",
+            "parameters": "4B",
+            "quantization": "Q4_K_M",
+            "size": "3.3 GB",
+            "description": "Vision-Language model, can analyze images/video inputs.",
+            "abilities": ["Vision", "Multimodal"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen3-vl",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "5.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 5.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "5.5 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 5.5
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "6.5 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 6.5
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "8.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 8.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "12.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 12.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "20.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 20.5
+                }
+            ],
+            "default_context": 4096
+        },
+        {
+            "id": "qwen3-vl:8b",
+            "name": "Qwen 3 VL 8B",
+            "creator": "Alibaba Cloud",
+            "parameters": "8B",
+            "quantization": "Q4_K_M",
+            "size": "6.1 GB",
+            "description": "Larger Vision-Language model for detailed visual reasoning.",
+            "abilities": ["Vision", "Multimodal", "Reasoning"],
+            "tool_support": true,
+            "reasoning": false,
+            "ollama_url": "https://ollama.com/library/qwen3-vl",
+            "max_context_window": 131072,
+            "context_profiles": [
+                {
+                    "size": 4096,
+                    "size_label": "4k",
+                    "vram_estimate": "8.0 GB",
+                    "ollama_context_size": 3482,
+                    "vram_estimate_gb": 8.0
+                },
+                {
+                    "size": 8192,
+                    "size_label": "8k",
+                    "vram_estimate": "8.7 GB",
+                    "ollama_context_size": 6963,
+                    "vram_estimate_gb": 8.7
+                },
+                {
+                    "size": 16384,
+                    "size_label": "16k",
+                    "vram_estimate": "10.0 GB",
+                    "ollama_context_size": 13926,
+                    "vram_estimate_gb": 10.0
+                },
+                {
+                    "size": 32768,
+                    "size_label": "32k",
+                    "vram_estimate": "12.5 GB",
+                    "ollama_context_size": 27853,
+                    "vram_estimate_gb": 12.5
+                },
+                {
+                    "size": 65536,
+                    "size_label": "64k",
+                    "vram_estimate": "17.5 GB",
+                    "ollama_context_size": 55706,
+                    "vram_estimate_gb": 17.5
+                },
+                {
+                    "size": 131072,
+                    "size_label": "128k",
+                    "vram_estimate": "27.5 GB",
+                    "ollama_context_size": 111411,
+                    "vram_estimate_gb": 27.5
+                }
+            ],
+            "default_context": 4096
+        }
+    ],
+    "fallback_model": {
+        "description": "No metadata available.",
+        "tool_support": false,
+        "context_sizes": [2048, 4096, 8192, 16384, 32768],
+        "default_context": 4096
+    }
+}