openbmb/minicpm-v4:q4_0

openbmb/

minicpm-v4:q4_0

408 Downloads Updated 3 weeks ago

A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone

vision 4b

Updated 1 month ago

1 month ago

bbf926ed13a0 · 3.0GB

archllama

·

parameters3.61B

·

quantizationQ4_0

2.1GB

archclip

·

parameters466M

·

quantizationF16

959MB

You are a helpful assistant.

28B

{ "num_ctx": 4096, "stop": [ "[\"<|im_start|>\",\"<|im_end|>\"]" ], "tempera

112B

{{- if .Messages }}{{- range $i, $_ := .Messages }}{{- $last := eq (len (slice $.Messages $i)) 1 -}}

488B

Readme

A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone

MiniCPM-V 4.0 🤗 🤖 | MiniCPM-o 2.6 🤗 🤖 | MiniCPM-V 2.6 🤗 🤖 | 🍳 Cookbook | 📄 Technical Blog [English/中文]

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:

🔥 Leading Visual Capability. 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.
🚀 Superior Efficiency. 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.
💫 Easy Usage. 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, featuring detailed instructions and practical examples.

<img src="/assets/openbmb/minicpm-v2.6/452bfaf1-53f4-4485-8a92-e4a1a05abf38" alt="MiniCPM-V.png" width="50%" style="display:block; margin:auto;">

**A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone**

<p align="center">
  MiniCPM-V 4.0 <a href="https://huggingface.co/openbmb/MiniCPM-V-4">🤗</a>  <a href="https://minicpm-v.openbmb.cn/"> 🤖</a> | MiniCPM-o 2.6 <a href="https://huggingface.co/openbmb/MiniCPM-o-2_6">🤗</a>  <a href="https://minicpm-omni-webdemo-us.modelbest.cn/"> 🤖</a> | MiniCPM-V 2.6 <a href="https://huggingface.co/openbmb/MiniCPM-V-2_6">🤗</a> <a href="http://120.92.209.146:8887/">🤖</a> | <a href="https://github.com/OpenSQZ/MiniCPM-V-CookBook">🍳 Cookbook</a> | 
  📄 Technical Blog [<a href="https://openbmb.notion.site/MiniCPM-o-2-6-A-GPT-4o-Level-MLLM-for-Vision-Speech-and-Multimodal-Live-Streaming-on-Your-Phone-185ede1b7a558042b5d5e45e6b237da9">English</a>/<a href="https://openbmb.notion.site/MiniCPM-o-2-6-GPT-4o-188ede1b7a558084b3aedd669cb80730">中文</a>] 
</p>

**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:

- 🔥 **Leading Visual Capability.**
   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.

- 🚀 **Superior Efficiency.**
  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.

-  💫  **Easy Usage.**
  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.

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)