{
  "catalogVersion": 1,
  "generatedAt": "2026-04-09T19:40:55Z",
  "families": [
    {
      "id": "llama",
      "name": "Llama",
      "tagline": "Meta's fast and capable open models",
      "symbolToken": "hare",
      "tintToken": "purple",
      "modelIDs": [
        "llama-3.2-1b-q4_k_m"
      ]
    },
    {
      "id": "gemma",
      "name": "Gemma",
      "tagline": "Google's lightweight open models",
      "symbolToken": "sparkles",
      "tintToken": "blue",
      "modelIDs": [
        "gemma-4-e2b-q4_k_m",
        "gemma-3-1b-q4_k_m"
      ]
    },
    {
      "id": "qwen",
      "name": "Qwen",
      "tagline": "Alibaba's efficient multilingual models",
      "symbolToken": "globe.asia.australia",
      "tintToken": "orange",
      "modelIDs": [
        "qwen3.5-0.8b-q4_k_m"
      ]
    },
    {
      "id": "phi",
      "name": "Phi",
      "tagline": "Microsoft's compact reasoning models",
      "symbolToken": "cpu",
      "tintToken": "indigo",
      "modelIDs": [
        "phi-3-mini-3.8b-q4_k_m"
      ]
    }
  ],
  "models": [
    {
      "id": "phi-3-mini-3.8b-q4_k_m",
      "name": "Phi-3 Mini 3.8B",
      "backend": "llamaCpp",
      "paramCount": "3.8B",
      "quantization": "Q4_K_M",
      "contextLength": 4096,
      "fileSizeBytes": 2393231360,
      "downloadURL": "https://huggingface.co/bartowski/Phi-3-mini-4k-instruct-GGUF/resolve/main/Phi-3-mini-4k-instruct-Q4_K_M.gguf",
      "description": "Compact general-purpose Phi model tuned for local llama.cpp inference."
    },
    {
      "id": "qwen3.5-0.8b-q4_k_m",
      "name": "Qwen3.5 0.8B",
      "backend": "llamaCpp",
      "paramCount": "0.8B",
      "quantization": "Q4_K_M",
      "contextLength": 32768,
      "fileSizeBytes": 558932992,
      "downloadURL": "https://huggingface.co/unsloth/Qwen3.5-0.8B-GGUF/resolve/main/Qwen3.5-0.8B-Q4_K_M.gguf",
      "description": "Ultra-compact Qwen3.5 model with a 32K context window."
    },
    {
      "id": "gemma-4-e2b-q4_k_m",
      "name": "Gemma 4 E2B",
      "backend": "llamaCpp",
      "paramCount": "2B",
      "quantization": "Q4_K_M",
      "contextLength": 131072,
      "fileSizeBytes": 3460000000,
      "downloadURL": "https://huggingface.co/bartowski/google_gemma-4-E2B-it-GGUF/resolve/main/google_gemma-4-E2B-it-Q4_K_M.gguf",
      "description": "Google's Gemma 4 efficient 2B model with 128K context and multimodal architecture."
    },
    {
      "id": "llama-3.2-1b-q4_k_m",
      "name": "Llama 3.2 1B",
      "backend": "llamaCpp",
      "paramCount": "1B",
      "quantization": "Q4_K_M",
      "contextLength": 131072,
      "fileSizeBytes": 762572800,
      "downloadURL": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/resolve/main/Llama-3.2-1B-Instruct-Q4_K_M.gguf",
      "description": "Meta's smallest Llama 3.2 model with a full 128K context window."
    },
    {
      "id": "gemma-3-1b-q4_k_m",
      "name": "Gemma 3 1B",
      "backend": "llamaCpp",
      "paramCount": "1B",
      "quantization": "Q4_K_M",
      "contextLength": 32768,
      "fileSizeBytes": 734003200,
      "downloadURL": "https://huggingface.co/bartowski/gemma-3-1b-it-GGUF/resolve/main/gemma-3-1b-it-Q4_K_M.gguf",
      "description": "Google's smallest Gemma 3 model with a 32K context window."
    }
  ]
}