whatmodel.sh / OpenThinker3 7B

OpenThinker3 7B

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

OpenThinker3 7B is an open-weights language model from Open Thoughts, released May 2025. Its published size is 7.62B parameters and a 33k-token context window. Licence: apache-2.0.

Catalogue identity notes

Maker repo open-thoughts/OpenThinker3-7B @864fb9a: licence tag apache-2.0. 7,615,616,512 parameters from the Hugging Face safetensors metadata. Dense, so active = total. Released 2025-05: month the maker's Hugging Face repo was created (createdAt 2025-05-27).

50.4±0.9#335 of 719 LLMsfrom 17 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of OpenThinker3 7B, out of 828 with a score, then more from Open Thoughts.

#modelscorewhereRAMreleased
402Qwen2 72B Instruct Alibaba51±1local~33GB2024-06
403Bielik 11B v2.0 Instruct SpeakLeash51±3local~8.5GB2024-08
404OpenThinker3 7B Open Thoughts50±1local~6GB2025-05
405CodeQwen1.5 7B Chat Alibaba50±1local~6.4GB2024-04
406Jamba 1.5 Large AI21 Labs50±1local~839GB2024-08
more from Open Thoughts
327OpenThinker 32B Open Thoughts55±1local~21GB2025-02
331OpenThinker2 32B Open Thoughts54±1local~21GB2025-04
428OpenThinker2 7B Open Thoughts49±2local~6.4GB2025-04

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #253 of 458 · 10% of its score · 3 tests

Beats Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B), GLM-4.7-Flash here, though they rank higher overall. Trails DeepSeek LLM 67B Chat, QwQ-32B-Preview, InternLM2.5 20B Chat here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEval+87.4% ✎▂·······▂·▃█ beats 2/6easy · avg 80%▇▇ 4.4%sharp
MBPP+61.4% ✎▄▄····▄·▄··█ beats 1/5medium · avg 62%▇ 3.3%at its level
LiveCodeBench v368.0% ✎▃▃·····▃·▃·█ beats 2/6easy · avg 65%▇ 2.9%at its level
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #338 of 393 * · 5.0% of its score · 2 tests

Trails Granite 4.0 H 350M, LFM2 350M, Gemma 3 1B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
AlpacaEval 2 length-controlled24.0% ✎█···█·█·█·██ beats 0/5medium · avg 44%▇▇ 4.4%at its level
IFEval (aggregation unspecified)51.7% ✎▄▄·····█▄▄·▄ beats 0/6easy · avg 79%▇ 0.6%noisy
math▁▂▃▅▆▅▄█▄▁▁▁ #167 of 398 · 37% of its score · 4 tests

Beats MiniMax-M2.1, Mistral Medium 3.5 (128B), GPT-4.1 here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
AIME 202467.7% ✎▄·····▄··▄██ beats 2/6medium · avg 59%▇▇▇▇▇ 13%sharp · at its level
AIME 202557.2% ✎▄··▄····█▄▄▄ beats 2/6medium · avg 51%▇▇▇▇▇ 11%sharp · at its level
MATH (subset unspecified)94.5% ✎▃·····▃···▅█ beats 3/6very easy · avg 90%▇▇▇ 7.7%typical
OMEGA (math generalisation)38.4% ✎▄······█▄▄·█ beats 3/6hard · avg 30%▇▇ 5.4%at its level
other tests8 not in a task · 47% of its score
testresultvs the others in its rundifficultycountswhy
BIG-Bench Hard77.1% ✎▄·····▄·▄·██ beats 2/6easy · avg 78%▇▇▇▇▇ 14%sharp · at its level
GPQA (subset unspecified)47.6% ✎▃······▃·▅·█ beats 3/6hard · avg 43%▇▇▇▇ 11%sharp · at its level
MMLU77.4% ✎▄···▄··█··▄█ beats 2/6easy · avg 77%▇▇▇ 7.0%at its level
AGIEval English78.6% ✎▄····▄·▄▄··█ beats 1/5easy · avg 76%▇▇ 5.1%typical
BRIDGE clinical text, zero-shot average (87 tasks)19.8% ✎▁▁▁▁▃▂▂▃▄▆█▃ beats 8/97hard · avg 32%▇ 3.7%blunt · at its level
ZebraLogic34.9% ✎█▄····▄·▄·▄▄ beats 2/6medium · avg 53%▇ 3.1%at its level
PopQA18.0% ✎▄··█···▄···█ beats 2/5hard · avg 22%▇ 2.1%blunt
IFBench (aggregation unspecified)23.0% ✎█·██·██····█ beats 0/5medium · avg 36%▇ 1.8%at its level

$ 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 OpenThinker3 7B · needs ~4.7 GB · from huggingface.co/lmstudio-community
    chat in the terminal
    llama cli -hf lmstudio-community/OpenThinker3-7B-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf lmstudio-community/OpenThinker3-7B-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 OpenThinker3 7B · needs ~4.7 GB · from huggingface.co/lmstudio-community

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/lmstudio-community/OpenThinker3-7B-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 OpenThinker3 7B · needs ~4.7 GB · from huggingface.co/lmstudio-community
    chat in the terminal
    ollama run hf.co/lmstudio-community/OpenThinker3-7B-GGUF:Q4_K_M

MLX (Mac)

version

  1. Install MLX (Mac) · Apple silicon Macs only, text only.
    with uv
    uv tool install mlx-lm
    or
    pip install mlx-lm
  2. Run OpenThinker3 7B · needs ~4.3 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/OpenThinker3-7B-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/OpenThinker3-7B-4bit

Other ways to run it

  1. Run it as published · needs ~18.79 GB, NVIDIA only · from huggingface.co/open-thoughts (official)
    vllm serve open-thoughts/OpenThinker3-7B

Official weights and documentation: huggingface.co/open-thoughts ↗

$ how the score was worked out · Olmo 3 83%, allenai/Olmo-3-7B-Think model… 13%, BRIDGE Medical Leaderboard (H… 4%

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
Olmo 314▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 83%CC BY 4.0
allenai/Olmo-3-7B-Think model card2▇▇▇▇ 13%Apache-2.0
BRIDGE Medical Leaderboard (Harvard Medical School / Brigham and Women's YLab)1▇ 4%MIT
prior · pull toward the middle▇ 0%
weighted average= 50.4

$ 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