alcoft/qwen_qwen3-4b-thinking-2507-gguf overview

{
  "metadata": {},
  "card_data": {
    "base_model": [
      "Qwen/Qwen3-4B-Thinking-2507"
    ],
    "pipeline_tag": "text-generation",
    "license": "apache-2.0",
    "license_link": "https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507/blob/main/LICENSE",
    "frontmatter": {
      "base_model": [
        "Qwen/Qwen3-4B-Thinking-2507"
      ],
      "pipeline_tag": "text-generation",
      "license": "apache-2.0",
      "license_link": "https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507/blob/main/LICENSE"
    },
    "hero_image_url": "",
    "summary": "|Quant|Size|Description| |---|---|---| |Q2_K_XXS|1.38 GB|Not recommended for most people. Extremelly low quality.| |Q2_K_XS|1.38 GB|Not recommended for most people. Very low quality.| |Q2_K|1.38 GB|Not recommended for most people. Very low quality.| |Q2_K_L|1.64 GB|Not recommended for most people. Uses Q8_0 for output and embedding, and Q2_K for everything else. Very low quality.| |Q2_K_XL|1.98 GB|Not recommended for most people. Uses F16 for output and embedding, and Q2_K for everything else. Very low quality.| |Q3_K_XXS|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Very low quality.| |Q3_K_XS|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Very low quality.| |Q3_K_S|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Low quality.| |Q3_K_M|1.79 GB|Not recommended for most people. Low quality.| |Q3_K_L|1.94 GB|Not recommended for most people. Low quality.| |Q3_K_XL|2.17 GB|Not recommended for most people. Uses Q8_0 for output and embedding, and Q3_K_L for everything else. Low quality.| |Q3_K_XXL|2.51 GB|Not recommended for most people. Uses F16 for output and embedding, and Q3_K_L for everything else. Low quality.| |Q4_K_XS|2.13 GB|Lower quality than Q4_K_S.| |Q4_K_S|2.13 GB|Recommended. Slightly low quality.| |Q4_K_M|2.23 GB|Recommended. Decent quality for most use cases.| |Q4_K_L|2.41 GB|Recommended. Uses Q8_0 for output and embedding, and Q4_K_M for everything else. Decent quality.| |Q4_K_XL|2.75 GB|Recommended. Uses F16 for output and embedding, and Q4_K_M for everything else. Decent quality.| |Q5_K_XXS|2.58 GB|Lower quality than Q5_K_S.| |Q5_K_XS|2.58 GB|Lower quality than Q5_K_S.| |Q5_K_S|2.58 GB|Recommended. High quality.| |Q5_K_M|2.64 GB|Recommended. High quality.| |Q5_K_L|2.78 GB|Recommended. Uses Q8_0 for output and embedding, and Q5_K_M for everything else. High quality.| |Q5_K_XL|3.12 GB|Recommended. Uses F16 for output and embedding, and Q5_K_M for everything else. High quality.| |Q6_K_S|3.08 GB|Lower quality than Q6_K.| |Q6_K|3.08 GB|Recommended. Very high quality.| |Q6_K_L|3.17 GB|Recommended. Uses Q8_0 for output and embedding, and Q6_K for everything else. Very high quality.| |Q6_K_XL|3.51 GB|Recommended. Uses F16 for output and embedding, and Q6_K for everything else. Very high quality.| |Q8_K_XS|3.99 GB|Lower quality than Q8_0.| |Q8_K_S|3.99 GB|Lower quality than Q8_0.| |Q8_0|3.99 GB|Recommended. Quality almost like F16.| |Q8_K_XL|4.33 GB|Recommended. Uses F16 for output and embedding, and Q8_0 for everything else. Quality almost like F16.| |F16|7.5 GB|Not recommended. Overkill. Prefer Q8_0.| |ORIGINAL (BF16)|7.5 GB|Not recommended. Overkill. Prefer Q8_0.| --- Quantized using TAO71-AI AutoQuantizer. You can check out the original model card here.",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "---\nbase_model:\n- Qwen/Qwen3-4B-Thinking-2507\npipeline_tag: text-generation\nlicense: apache-2.0\nlicense_link: https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507/blob/main/LICENSE\n---\n\n|Quant|Size|Description|\n|---|---|---|\n|[Q2_K_XXS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q2_K_XXS.gguf)|1.38 GB|Not recommended for most people. Extremelly low quality.|\n|[Q2_K_XS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q2_K_XS.gguf)|1.38 GB|Not recommended for most people. Very low quality.|\n|[Q2_K](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q2_K.gguf)|1.38 GB|Not recommended for most people. Very low quality.|\n|[Q2_K_L](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q2_K_L.gguf)|1.64 GB|Not recommended for most people. Uses Q8_0 for output and embedding, and Q2_K for everything else. Very low quality.|\n|[Q2_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q2_K_XL.gguf)|1.98 GB|Not recommended for most people. Uses F16 for output and embedding, and Q2_K for everything else. Very low quality.|\n|[Q3_K_XXS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_XXS.gguf)|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Very low quality.|\n|[Q3_K_XS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_XS.gguf)|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Very low quality.|\n|[Q3_K_S](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_S.gguf)|1.62 GB|Not recommended for most people. Prefer any bigger Q3_K quantization. Low quality.|\n|[Q3_K_M](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_M.gguf)|1.79 GB|Not recommended for most people. Low quality.|\n|[Q3_K_L](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_L.gguf)|1.94 GB|Not recommended for most people. Low quality.|\n|[Q3_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_XL.gguf)|2.17 GB|Not recommended for most people. Uses Q8_0 for output and embedding, and Q3_K_L for everything else. Low quality.|\n|[Q3_K_XXL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q3_K_XXL.gguf)|2.51 GB|Not recommended for most people. Uses F16 for output and embedding, and Q3_K_L for everything else. Low quality.|\n|[Q4_K_XS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q4_K_XS.gguf)|2.13 GB|Lower quality than Q4_K_S.|\n|[Q4_K_S](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q4_K_S.gguf)|2.13 GB|Recommended. Slightly low quality.|\n|[Q4_K_M](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q4_K_M.gguf)|2.23 GB|Recommended. Decent quality for most use cases.|\n|[Q4_K_L](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q4_K_L.gguf)|2.41 GB|Recommended. Uses Q8_0 for output and embedding, and Q4_K_M for everything else. Decent quality.|\n|[Q4_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q4_K_XL.gguf)|2.75 GB|Recommended. Uses F16 for output and embedding, and Q4_K_M for everything else. Decent quality.|\n|[Q5_K_XXS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_XXS.gguf)|2.58 GB|Lower quality than Q5_K_S.|\n|[Q5_K_XS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_XS.gguf)|2.58 GB|Lower quality than Q5_K_S.|\n|[Q5_K_S](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_S.gguf)|2.58 GB|Recommended. High quality.|\n|[Q5_K_M](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_M.gguf)|2.64 GB|Recommended. High quality.|\n|[Q5_K_L](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_L.gguf)|2.78 GB|Recommended. Uses Q8_0 for output and embedding, and Q5_K_M for everything else. High quality.|\n|[Q5_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q5_K_XL.gguf)|3.12 GB|Recommended. Uses F16 for output and embedding, and Q5_K_M for everything else. High quality.|\n|[Q6_K_S](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q6_K_S.gguf)|3.08 GB|Lower quality than Q6_K.|\n|[Q6_K](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q6_K.gguf)|3.08 GB|Recommended. Very high quality.|\n|[Q6_K_L](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q6_K_L.gguf)|3.17 GB|Recommended. Uses Q8_0 for output and embedding, and Q6_K for everything else. Very high quality.|\n|[Q6_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q6_K_XL.gguf)|3.51 GB|Recommended. Uses F16 for output and embedding, and Q6_K for everything else. Very high quality.|\n|[Q8_K_XS](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q8_K_XS.gguf)|3.99 GB|Lower quality than Q8_0.|\n|[Q8_K_S](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q8_K_S.gguf)|3.99 GB|Lower quality than Q8_0.|\n|[Q8_0](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q8_0.gguf)|3.99 GB|Recommended. Quality almost like F16.|\n|[Q8_K_XL](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_Q8_K_XL.gguf)|4.33 GB|Recommended. Uses F16 for output and embedding, and Q8_0 for everything else. Quality almost like F16.|\n|[F16](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507_F16.gguf)|7.5 GB|Not recommended. Overkill. Prefer Q8_0.|\n|[ORIGINAL (BF16)](https://huggingface.co/Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF/resolve/main/Qwen_Qwen3-4B-Thinking-2507.gguf)|7.5 GB|Not recommended. Overkill. Prefer Q8_0.|\n\n---\n\nQuantized using [TAO71-AI AutoQuantizer](https://github.com/TAO71-AI/AutoQuantizer).\nYou can check out the original model card [here](https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507).",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "text-generation",
    "base_model:Qwen/Qwen3-4B-Thinking-2507",
    "base_model:quantized:Qwen/Qwen3-4B-Thinking-2507",
    "license:apache-2.0",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 329,
  "gated": false,
  "private": false,
  "last_modified": "2025-10-22T15:20:58.000Z",
  "created_at": "2025-10-22T15:00:31.000Z",
  "pipeline_tag": "text-generation",
  "library_name": ""
}

{
  "_id": "68f8f18fc772e39184c2aa58",
  "id": "Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF",
  "modelId": "Alcoft/Qwen_Qwen3-4B-Thinking-2507-GGUF",
  "sha": "8c8fc0ea62e77cb4fe4747599f5ee425cfcd41ad",
  "createdAt": "2025-10-22T15:00:31.000Z",
  "lastModified": "2025-10-22T15:20:58.000Z",
  "author": "Alcoft",
  "downloads": 329,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "text-generation",
  "library_name": "",
  "siblings_count": 35
}