openbmb/minicpm-v-4-gguf overview

{
  "metadata": {},
  "card_data": {
    "pipeline_tag": "image-text-to-text",
    "datasets": [
      "openbmb/RLAIF-V-Dataset"
    ],
    "library_name": "transformers",
    "language": [
      "multilingual"
    ],
    "tags": [
      "minicpm-v",
      "vision",
      "ocr",
      "multi-image",
      "video",
      "custom_code"
    ],
    "frontmatter": {
      "pipeline_tag": "image-text-to-text",
      "datasets": [
        "openbmb/RLAIF-V-Dataset"
      ],
      "library_name": "transformers",
      "language": [
        "multilingual"
      ],
      "tags": [
        "minicpm-v",
        "vision",
        "ocr",
        "multi-image",
        "video",
        "custom_code"
      ]
    },
    "hero_image_url": "https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/minicpm-v-4-case.png",
    "summary": "question = \"What is the landform in the picture?\" msgs = [{'role': 'user', 'content': [image, question]}] answer = model.chat( msgs=msgs, image=image, tokenizer=tokenizer ) print(answer) # Second round chat, pass history context of multi-turn conversation msgs.append({\"role\": \"assistant\", \"content\": [answer]}) msgs.append({\"role\": \"user\", \"content\": [ \"What should I pay attention to when traveling here?\"]}) answer = model.chat( msgs=msgs, image=None, tokenizer=tokenizer ) print(answer) ```",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "---\npipeline_tag: image-text-to-text\ndatasets:\n- openbmb/RLAIF-V-Dataset\nlibrary_name: transformers\nlanguage:\n- multilingual\ntags:\n- minicpm-v\n- vision\n- ocr\n- multi-image\n- video\n- custom_code\n---\n\n<h1>A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone</h1>\n\n[GitHub](https://github.com/OpenBMB/MiniCPM-o) | [Demo](http://211.93.21.133:8889/)</a> \n\n\n\n## MiniCPM-V 4.0\n\n**MiniCPM-V 4.0** is the latest efficient model in the MiniCPM-V series. The model is built based on SigLIP2-400M and MiniCPM4-3B with a total of 4.1B parameters. It inherits the strong single-image, multi-image and video understanding performance of MiniCPM-V 2.6 with largely improved efficiency. Notable features of MiniCPM-V 4.0 include:\n\n- 🔥 **Leading Visual Capability.**\n   With only 4.1B parameters, MiniCPM-V 4.0 achieves an average score of 69.0 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks, **outperforming GPT-4.1-mini-20250414, MiniCPM-V 2.6 (8.1B params, OpenCompass 65.2) and Qwen2.5-VL-3B-Instruct (3.8B params, OpenCompass 64.5)**. It also shows good performance in multi-image understanding and video understanding.\n\n- 🚀 **Superior Efficiency.**\n  Designed for on-device deployment, MiniCPM-V 4.0 runs smoothly on end devices. For example, it devlivers **less than 2s first token delay and more than 17 token/s decoding on iPhone 16 Pro Max**, without heating problems. It also shows superior throughput under concurrent requests.\n\n-  💫  **Easy Usage.**\n  MiniCPM-V 4.0 can be easily used in various ways including **llama.cpp, Ollama, vLLM, SGLang, LLaMA-Factory and local web demo** etc. We also open-source iOS App that can run on iPhone and iPad. Get started easily with our well-structured [Cookbook](https://github.com/OpenSQZ/MiniCPM-V-CookBook), featuring detailed instructions and practical examples.\n\n\n### Evaluation\n\n<details>\n<summary>Click to view single image results on OpenCompass. </summary>\n<div align=\"center\">\n<table style=\"margin: 0px auto;\">\n    <thead>\n        <tr>\n            <th nowrap=\"nowrap\" align=\"left\">model</th>\n            <th>Size</th>\n            <th>Opencompass</th>\n            <th>OCRBench</th>\n            <th>MathVista</th>\n            <th>HallusionBench</th>\n            <th>MMMU</th>\n            <th>MMVet</th>\n            <th>MMBench V1.1</th>\n            <th>MMStar</th>\n            <th>AI2D</th>\n        </tr>\n    </thead>\n    <tbody align=\"center\">\n        <tr>\n            <td colspan=\"11\" align=\"left\"><strong>Proprietary</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4v-20240409</td>\n            <td>-</td>\n            <td>63.5</td>\n            <td>656</td>\n            <td>55.2</td>\n            <td>43.9</td>\n            <td>61.7</td>\n            <td>67.5</td>\n            <td>79.8</td>\n            <td>56.0</td>\n            <td>78.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Gemini-1.5-Pro</td>\n            <td>-</td>\n            <td>64.5</td>\n            <td>754</td>\n            <td>58.3</td>\n            <td>45.6</td>\n            <td>60.6</td>\n            <td>64.0</td>\n            <td>73.9</td>\n            <td>59.1</td>\n            <td>79.1</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4.1-mini-20250414</td>\n            <td>-</td>\n            <td>68.9</td>\n            <td>840</td>\n            <td>70.9</td>\n            <td>49.3</td>\n            <td>55.0</td>\n            <td>74.3</td>\n            <td>80.9</td>\n            <td>60.9</td>\n            <td>76.0</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Claude 3.5 Sonnet-20241022</td>\n            <td>-</td>\n            <td>70.6</td>\n            <td>798</td>\n            <td>65.3</td>\n            <td>55.5</td>\n            <td>66.4</td>\n            <td>70.1</td>\n            <td>81.7</td>\n            <td>65.1</td>\n            <td>81.2</td>\n        </tr>\n        <tr>\n            <td colspan=\"11\" align=\"left\"><strong>Open-source</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-3B-Instruct</td>\n            <td>3.8B</td>\n            <td>64.5</td>\n            <td>828</td>\n            <td>61.2</td>\n            <td>46.6</td>\n            <td>51.2</td>\n            <td>60.0</td>\n            <td>76.8</td>\n            <td>56.3</td>\n            <td>81.4</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-4B</td>\n            <td>3.7B</td>\n            <td>65.1</td>\n            <td>820</td>\n            <td>60.8</td>\n            <td>46.6</td>\n            <td>51.8</td>\n            <td>61.5</td>\n            <td>78.2</td>\n            <td>58.7</td>\n            <td>81.4</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-7B-Instruct</td>\n            <td>8.3B</td>\n            <td>70.9</td>\n            <td>888</td>\n            <td>68.1</td>\n            <td>51.9</td>\n            <td>58.0</td>\n            <td>69.7</td>\n            <td>82.2</td>\n            <td>64.1</td>\n            <td>84.3</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-8B</td>\n            <td>8.1B</td>\n            <td>68.1</td>\n            <td>821</td>\n            <td>64.5</td>\n            <td>49.0</td>\n            <td>56.2</td>\n            <td>62.8</td>\n            <td>82.5</td>\n            <td>63.2</td>\n            <td>84.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-2.6</td>\n            <td>8.1B</td>\n            <td>65.2</td>\n            <td>852</td>\n            <td>60.8</td>\n            <td>48.1</td>\n            <td>49.8</td>\n            <td>60.0</td>\n            <td>78.0</td>\n            <td>57.5</td>\n            <td>82.1</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-o-2.6</td>\n            <td>8.7B</td>\n            <td>70.2</td>\n            <td>889</td>\n            <td>73.3</td>\n            <td>51.1</td>\n            <td>50.9</td>\n            <td>67.2</td>\n            <td>80.6</td>\n            <td>63.3</td>\n            <td>86.1</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-4.0</td>\n            <td>4.1B</td>\n            <td>69.0</td>\n            <td>894</td>\n            <td>66.9</td>\n            <td>50.8</td>\n            <td>51.2</td>\n            <td>68.0</td>\n            <td>79.7</td>\n            <td>62.8</td>\n            <td>82.9</td>\n        </tr>\n    </tbody>\n</table>\n</div>\n\n</details>\n\n<details>\n<summary>Click to view single image results on ChartQA, MME, RealWorldQA, TextVQA, DocVQA, MathVision, DynaMath, WeMath, Object HalBench and MM Halbench. </summary>\n\n<div align=\"center\">\n<table style=\"margin: 0px auto;\">\n    <thead>\n        <tr>\n            <th nowrap=\"nowrap\" align=\"left\">model</th>\n            <th>Size</th>\n            <th>ChartQA</th>\n            <th>MME</th>\n            <th>RealWorldQA</th>\n            <th>TextVQA</th>\n            <th>DocVQA</th>\n            <th>MathVision</th>\n            <th>DynaMath</th>\n            <th>WeMath</th>\n            <th colspan=\"2\">Obj Hal</th>\n            <th colspan=\"2\">MM Hal</th>\n        </tr>\n    </thead>\n    <tbody>\n        <tr>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td>CHAIRs↓</td>\n            <td>CHAIRi↓</td>\n            <td nowrap=\"nowrap\">score avg@3↑</td>\n            <td nowrap=\"nowrap\">hall rate avg@3↓</td>\n        </tr>\n        <tbody align=\"center\">\n        <tr>\n            <td colspan=\"14\" align=\"left\"><strong>Proprietary</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4v-20240409</td>\n            <td>-</td>\n            <td>78.5</td>\n            <td>1927</td>\n            <td>61.4</td>\n            <td>78.0</td>\n            <td>88.4</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Gemini-1.5-Pro</td>\n            <td>-</td>\n            <td>87.2</td>\n            <td>-</td>\n            <td>67.5</td>\n            <td>78.8</td>\n            <td>93.1</td>\n            <td>41.0</td>\n            <td>31.5</td>\n            <td>50.5</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4.1-mini-20250414</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>45.3</td>\n            <td>47.7</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Claude 3.5 Sonnet-20241022</td>\n            <td>-</td>\n            <td>90.8</td>\n            <td>-</td>\n            <td>60.1</td>\n            <td>74.1</td>\n            <td>95.2</td>\n            <td>35.6</td>\n            <td>35.7</td>\n            <td>44.0</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n            <td>-</td>\n        </tr>\n        <tr>\n            <td colspan=\"14\" align=\"left\"><strong>Open-source</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-3B-Instruct</td>\n            <td>3.8B</td>\n            <td>84.0</td>\n            <td>2157</td>\n            <td>65.4</td>\n            <td>79.3</td>\n            <td>93.9</td>\n            <td>21.9</td>\n            <td>13.2</td>\n            <td>22.9</td>\n            <td>18.3</td>\n            <td>10.8</td>\n            <td>3.9 </td>\n            <td>33.3 </td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-4B</td>\n            <td>3.7B</td>\n            <td>84.0</td>\n            <td>2338</td>\n            <td>64.3</td>\n            <td>76.8</td>\n            <td>91.6</td>\n            <td>18.4</td>\n            <td>15.2</td>\n            <td>21.2</td>\n            <td>13.7</td>\n            <td>8.7</td>\n            <td>3.2 </td>\n            <td>46.5 </td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-7B-Instruct</td>\n            <td>8.3B</td>\n            <td>87.3</td>\n            <td>2347</td>\n            <td>68.5</td>\n            <td>84.9</td>\n            <td>95.7</td>\n            <td>25.4</td>\n            <td>21.8</td>\n            <td>36.2</td>\n            <td>13.3</td>\n            <td>7.9</td>\n            <td>4.1 </td>\n            <td>31.6 </td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-8B</td>\n            <td>8.1B</td>\n            <td>84.8</td>\n            <td>2344</td>\n            <td>70.1</td>\n            <td>79.1</td>\n            <td>93.0</td>\n            <td>17.0</td>\n            <td>9.4</td>\n            <td>23.5</td>\n            <td>18.3</td>\n            <td>11.6</td>\n            <td>3.6 </td>\n            <td>37.2</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-2.6</td>\n            <td>8.1B</td>\n            <td>79.4</td>\n            <td>2348</td>\n            <td>65.0</td>\n            <td>80.1</td>\n            <td>90.8</td>\n            <td>17.5</td>\n            <td>9.0</td>\n            <td>20.4</td>\n            <td>7.3</td>\n            <td>4.7</td>\n            <td>4.0 </td>\n            <td>29.9 </td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-o-2.6</td>\n            <td>8.7B</td>\n            <td>86.9</td>\n            <td>2372</td>\n            <td>68.1</td>\n            <td>82.0</td>\n            <td>93.5</td>\n            <td>21.7</td>\n            <td>10.4</td>\n            <td>25.2</td>\n            <td>6.3</td>\n            <td>3.4</td>\n            <td>4.1 </td>\n            <td>31.3 </td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-4.0</td>\n            <td>4.1B</td>\n            <td>84.4</td>\n            <td>2298</td>\n            <td>68.5</td>\n            <td>80.8</td>\n            <td>92.9</td>\n            <td>20.7</td>\n            <td>14.2</td>\n            <td>32.7</td>\n            <td>6.3</td>\n            <td>3.5</td>\n            <td>4.1 </td>\n            <td>29.2 </td>\n        </tr>\n    </tbody>\n</table>\n</div>\n\n</details>\n\n<details>\n<summary>Click to view multi-image and video understanding results on Mantis, Blink and Video-MME. </summary>\n<div align=\"center\">\n<table style=\"margin: 0px auto;\">\n    <thead>\n        <tr>\n            <th nowrap=\"nowrap\" align=\"left\">model</th>\n            <th>Size</th>\n            <th>Mantis</th>\n            <th>Blink</th>\n            <th nowrap=\"nowrap\" colspan=\"2\" >Video-MME</th>\n        </tr>\n    </thead>\n    <tbody>\n        <tr>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td></td>\n            <td>wo subs</td>\n            <td>w subs</td>\n        </tr>\n        <tbody align=\"center\">\n        <tr>\n            <td colspan=\"6\" align=\"left\"><strong>Proprietary</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4v-20240409</td>\n            <td>-</td>\n            <td>62.7</td>\n            <td>54.6</td>\n            <td>59.9</td>\n            <td>63.3</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Gemini-1.5-Pro</td>\n            <td>-</td>\n            <td>-</td>\n            <td>59.1</td>\n            <td>75.0</td>\n            <td>81.3</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">GPT-4o-20240513</td>\n            <td>-</td>\n            <td>-</td>\n            <td>68.0</td>\n            <td>71.9</td>\n            <td>77.2</td>\n        </tr>\n        <tr>\n            <td colspan=\"6\" align=\"left\"><strong>Open-source</strong></td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-3B-Instruct</td>\n            <td>3.8B</td>\n            <td>-</td>\n            <td>47.6</td>\n            <td>61.5</td>\n            <td>67.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-4B</td>\n            <td>3.7B</td>\n            <td>62.7</td>\n            <td>50.8</td>\n            <td>62.3</td>\n            <td>63.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">Qwen2.5-VL-7B-Instruct</td>\n            <td>8.3B</td>\n            <td>-</td>\n            <td>56.4</td>\n            <td>65.1</td>\n            <td>71.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">InternVL2.5-8B</td>\n            <td>8.1B</td>\n            <td>67.7</td>\n            <td>54.8</td>\n            <td>64.2</td>\n            <td>66.9</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-2.6</td>\n            <td>8.1B</td>\n            <td>69.1</td>\n            <td>53.0</td>\n            <td>60.9</td>\n            <td>63.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-o-2.6</td>\n            <td>8.7B</td>\n            <td>71.9</td>\n            <td>56.7</td>\n            <td>63.9</td>\n            <td>69.6</td>\n        </tr>\n        <tr>\n            <td nowrap=\"nowrap\" align=\"left\">MiniCPM-V-4.0</td>\n            <td>4.1B</td>\n            <td>71.4</td>\n            <td>54.0</td>\n            <td>61.2</td>\n            <td>65.8</td>\n        </tr>\n    </tbody>\n</table>\n</div>\n\n</details>\n\n### Examples\n\n<div style=\"display: flex; flex-direction: column; align-items: center;\">\n  <img src=\"https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/minicpm-v-4-case.png\" alt=\"math\" style=\"margin-bottom: 5px;\">\n</div>\n\nRun locally on iPhone 16 Pro Max with [iOS demo](https://github.com/OpenSQZ/MiniCPM-V-CookBook/blob/main/demo/ios_demo/ios.md).\n\n<div align=\"center\">\n  <img src=\"https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/iphone_en.gif\" width=\"45%\" style=\"display: inline-block; margin: 0 10px;\"/>\n  <img src=\"https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/iphone_en_information_extraction.gif\" width=\"45%\" style=\"display: inline-block; margin: 0 10px;\"/>\n</div>\n\n<div align=\"center\">\n  <img src=\"https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/iphone_cn.gif\" width=\"45%\" style=\"display: inline-block; margin: 0 10px;\"/>\n  <img src=\"https://raw.githubusercontent.com/openbmb/MiniCPM-o/main/assets/minicpmv4/iphone_cn_funny_points.gif\" width=\"45%\" style=\"display: inline-block; margin: 0 10px;\"/>\n</div> \n\n## Usage\n\n```python\nfrom PIL import Image\nimport torch\nfrom transformers import AutoModel, AutoTokenizer\n\nmodel_path = 'openbmb/MiniCPM-V-4'\nmodel = AutoModel.from_pretrained(model_path, trust_remote_code=True,\n                                  # sdpa or flash_attention_2, no eager\n                                  attn_implementation='sdpa', torch_dtype=torch.bfloat16)\nmodel = model.eval().cuda()\ntokenizer = AutoTokenizer.from_pretrained(\n    model_path, trust_remote_code=True)\n\n\n\nimage = Image.open('./assets/single.png').convert('RGB')\n\n# First round chat \nquestion = \"What is the landform in the picture?\"\nmsgs = [{'role': 'user', 'content': [image, question]}]\n\nanswer = model.chat(\n    msgs=msgs,\n    image=image,\n    tokenizer=tokenizer\n)\nprint(answer)\n\n\n# Second round chat, pass history context of multi-turn conversation\nmsgs.append({\"role\": \"assistant\", \"content\": [answer]})\nmsgs.append({\"role\": \"user\", \"content\": [\n            \"What should I pay attention to when traveling here?\"]})\n\nanswer = model.chat(\n    msgs=msgs,\n    image=None,\n    tokenizer=tokenizer\n)\nprint(answer)\n```\n\n\n## License\n#### Model License\n* The code in this repo is released under the [Apache-2.0](https://github.com/OpenBMB/MiniCPM/blob/main/LICENSE) License. \n* The usage of MiniCPM-V series model weights must strictly follow [MiniCPM Model License.md](https://github.com/OpenBMB/MiniCPM/blob/main/MiniCPM%20Model%20License.md).\n* The models and weights of MiniCPM are completely free for academic research. After filling out a [\"questionnaire\"](https://modelbest.feishu.cn/share/base/form/shrcnpV5ZT9EJ6xYjh3Kx0J6v8g) for registration, MiniCPM-V 2.6 weights are also available for free commercial use.\n\n\n#### Statement\n* As an LMM, MiniCPM-V 4.0 generates contents by learning a large mount of multimodal corpora, but it cannot comprehend, express personal opinions or make value judgement. Anything generated by MiniCPM-V 4.0 does not represent the views and positions of the model developers\n* We will not be liable for any problems arising from the use of the MinCPM-V models, including but not limited to data security issues, risk of public opinion, or any risks and problems arising from the misdirection, misuse, dissemination or misuse of the model.\n\n## Key Techniques and Other Multimodal Projects\n\n👏 Welcome to explore key techniques of MiniCPM-V 2.6 and other multimodal projects of our team:\n\n[VisCPM](https://github.com/OpenBMB/VisCPM/tree/main) | [RLHF-V](https://github.com/RLHF-V/RLHF-V) | [LLaVA-UHD](https://github.com/thunlp/LLaVA-UHD)  | [RLAIF-V](https://github.com/RLHF-V/RLAIF-V)\n\n## Citation\n\nIf you find our work helpful, please consider citing our papers 📝 and liking this project ❤️！\n\n```bib\n@article{yao2024minicpm,\n  title={MiniCPM-V: A GPT-4V Level MLLM on Your Phone},\n  author={Yao, Yuan and Yu, Tianyu and Zhang, Ao and Wang, Chongyi and Cui, Junbo and Zhu, Hongji and Cai, Tianchi and Li, Haoyu and Zhao, Weilin and He, Zhihui and others},\n  journal={Nat Commun 16, 5509 (2025)},\n  year={2025}\n}\n```\n\n",
    "related_quantizations": []
  },
  "tags": [
    "transformers",
    "gguf",
    "minicpm-v",
    "vision",
    "ocr",
    "multi-image",
    "video",
    "custom_code",
    "image-text-to-text",
    "multilingual",
    "dataset:openbmb/RLAIF-V-Dataset",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 48,
  "downloads": 1831,
  "gated": false,
  "private": false,
  "last_modified": "2025-09-19T07:48:53.000Z",
  "created_at": "2025-07-12T09:45:29.000Z",
  "pipeline_tag": "image-text-to-text",
  "library_name": "transformers"
}

{
  "_id": "68722eb903ab62ee6f8a46e3",
  "id": "openbmb/MiniCPM-V-4-gguf",
  "modelId": "openbmb/MiniCPM-V-4-gguf",
  "sha": "8f4a0cfdd5960b4b3b7793600b2d42533dcce9ca",
  "createdAt": "2025-07-12T09:45:29.000Z",
  "lastModified": "2025-09-19T07:48:53.000Z",
  "author": "openbmb",
  "downloads": 1831,
  "likes": 48,
  "gated": false,
  "private": false,
  "pipeline_tag": "image-text-to-text",
  "library_name": "transformers",
  "siblings_count": 24
}