whatmodel.sh / MiniCPM-V 2.6

MiniCPM-V 2.6

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

MiniCPM-V 2.6 is an open-weights language model from OpenBMB, released August 2024. Its published size is 8.1B parameters. Licence: minicpm-model-license.

Catalogue identity notes

Arena entry 'minicpm-v-2_6' (vision). LMArena's leaderboard table (arena-leaderboard space @6f2eeeb) labels 'minicpm-v-2_6' as 'MiniCPM-v 2_6' (OpenBMB, Apache 2.0), linking https://huggingface.co/openbmb/MiniCPM-V-2_6. Maker repo openbmb/MiniCPM-V-2_6 @6c04d9e: licence tag none tagged. 8,099,175,152 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2024-08: month the maker's Hugging Face repo was created (createdAt 2024-08-04).

Model information sources

· Weights licence ↗

42.8±1.9#484 of 719 LLMsfrom 5 benchmark results across 4 sources

$ how it compares

The LLMs ranked either side of MiniCPM-V 2.6, out of 828 with a score, then more from OpenBMB.

#modelscorewhereRAMreleased
572Reka Flash 21B (2024-02-26) Reka AI43±1comparison only2024-02
573T5 Small Google43±9local~1.8GB2019-10
574MiniCPM-V 2.6 OpenBMB43±2local~6.4GB2024-08
575Qwen2.5 Coder 1.5B Instruct Alibaba43±1local~2.4GB2024-09
576LFM2 2.6B Liquid AI43±2local~3GB2025-09
more from OpenBMB
186MiniCPM5 2B OpenBMB64±1local~3GB2026-09
467MiniCPM3-4B OpenBMB48±2local~4GB2024-09

see all LLMs · find the ones that fit your machine

$ what it's good at

vision & documents▁▃▃▂▁▃▄▂▆▇█▄ #141 of 147 · 98% of its score · 5 tests

Beats Gemma 3 4B here, though they rank higher overall. Trails Qwen3.5 0.8B, Molmo 7B-D (0924), North Micro Vision Instruct (2.4B) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
CharXiv descriptive questions57.0% ✎▄█▄▄▄·██·▄·█ beats 2/12easy · avg 71%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 42%at its level
Arena vision (style controlled)962.7 ✎▁▁▃▂▂▂▄▅▅▆██ beats 3/145avg 1068▇▇▇▇▇▇▇▇▇ 21%blunt
CharXiv (reasoning questions)31.0% ✎▂▂▂█▃▂·▂▂··▂ beats 3/13medium · avg 36%▇▇▇▇▇▇▇ 16%typical
CapArena-Auto detailed image captioning (Score_Avg, GPT-4o judge)-38.0000 ✎▃▃▃·▃█·▅▃▃▃▃ beats 2/12avg -0.671▇▇▇▇▇▇ 15%typical
PhysBench physical-world understanding (test, Avg)44.9% ✎▅▃█··▅▃····▃ beats 8/9hard · avg 43%▇ 3.7%blunt

$ where to run it

llama.cpp · recommended

version

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run MiniCPM-V 2.6 · needs ~4.7 GB · from huggingface.co/openbmb (official)
    chat in the terminal
    llama cli -hf openbmb/MiniCPM-V-2_6-gguf:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf openbmb/MiniCPM-V-2_6-gguf:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run MiniCPM-V 2.6 · needs ~4.7 GB · from huggingface.co/openbmb (official)

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/openbmb/MiniCPM-V-2_6-gguf@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run MiniCPM-V 2.6 · needs ~5.36 GB · from ollama.com
    chat in the terminal
    ollama run minicpm-v:8b-2.6-q4_K_M

Official weights and documentation: huggingface.co/openbmb ↗

$ how the score was worked out · CharXiv leaderboard (princeto… 59%, LMArena leaderboard dataset 21%, CapArena: Benchmarking and An… 15% and 1 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
CharXiv leaderboard (princeton-nlp/CharXiv README)2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 59%Apache-2.0
LMArena leaderboard dataset1▇▇▇▇▇▇ 21%CC BY 4.0
CapArena: Benchmarking and Analyzing Detailed Image Captioning in the LLM Era (arXiv:2503.12329v1)1▇▇▇▇ 15%CC-BY-4.0
PhysBench: Benchmarking and Enhancing Vision-Language Models for Physical World Understanding (arXiv:2501.16411v2)1▇ 4%CC-BY-4.0
prior · pull toward the middle▇ 2%
weighted average= 42.8

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json