whatmodel.sh / Olmo 3 7B Think

Olmo 3 7B Think

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

Olmo 3 7B Think is an open-weights language model from Ai2, released November 2025. Its published size is 7.3B parameters and a 66k-token context window. Licence: apache-2.0.

52.9±0.8#301 of 719 LLMsfrom 25 benchmark results across 5 sources

$ how it compares

The LLMs ranked either side of Olmo 3 7B Think, out of 828 with a score, then more from Ai2.

#modelscorewhereRAMreleased
359s1.1 32B Simple Scaling53±1local~22GB2025-02
360DeepSeek-Coder-V2 Instruct (0724) DeepSeek53±1local~151GB2024-09
361Olmo 3 7B Think Ai253±1local~5.8GB2025-11
362GPT-4.1 nano OpenAI53±1api $0.17/M2025-04
363Gemini 1.5 Flash (002) Google53±1comparison only2024-09
more from Ai2
259Olmo 3.1 32B Think Ai258±1local~21GB2025-12
262Olmo 3 32B Think Ai258±1local~21GB2025-11
314Molmo 2 8B Ai255±2local~6.8GB2025-12

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #230 of 458 · 10% of its score · 4 tests

Beats Qwen3.5 4B, Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B) here, though they rank higher overall. Trails QwQ-32B-Preview, Ministral 3 3B Reasoning 2512, GPT-4 (0613) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEval+ lab89.9% ✎▂·······▂·▂█ beats 4/6easy · avg 80%▇ 3.0%sharp
MBPP+ lab64.7% ✎▄▄····▄·▄··█ beats 2/5medium · avg 62%▇ 2.6%at its level
LiveCodeBench v6 (window unspecified)59.8% ✎▂▃▄▄▅██▅▄▇▃▆ beats 24/65medium · avg 43%▇ 2.2%at its level
LiveCodeBench v3 lab75.2% ✎▃▃·····▃·▃·█ beats 3/6easy · avg 65%▇ 2.2%at its level
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #197 of 393 · 4.7% of its score · 2 tests

Beats Granite 4.2 8B, Qwen3-VL 30B-A3B Thinking, Mistral Small 4 (119B-A6.5B) here, though they rank higher overall. Trails LFM2.5 1.2B Instruct, LFM2.5 1.2B Thinking, Gemma 3 12B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
AlpacaEval 2 length-controlled lab52.1% ✎█···█·█·█·██ beats 2/5medium · avg 44%▇ 3.6%at its level
IFEval (aggregation unspecified) lab +188.2% ✎▄▄·····█▄▄·▄ beats 4/6easy · avg 76%▇ 1.1%noisy · several runs
math▁▂▃▅▆▅▄█▄▁▁▁ #157 of 398 · 29% of its score · 4 tests

Beats MiMo-V2.5 (310B-A15B), MiniMax-M2.7, MiniMax-M2.1 here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
AIME 2024 lab71.6% ✎▄·····▄··▄██ beats 3/6medium · avg 59%▇▇▇▇ 10%sharp · at its level
AIME 2025 lab64.6% ✎▄··▄····█▄▄▄ beats 4/6medium · avg 51%▇▇▇▇ 9.0%sharp · at its level
MATH (subset unspecified) lab95.1% ✎▃·····▃···▅█ beats 5/6very easy · avg 90%▇▇ 5.3%maxed out
OMEGA (math generalisation) lab37.8% ✎▄······█▄▄·█ beats 1/6hard · avg 30%▇▇ 4.7%at its level
other tests14 not in a task · 56% of its score
testresultvs the others in its rundifficultycountswhy
BIG-Bench Hard lab86.6% ✎▄·····▄·▄·██ beats 4/6easy · avg 78%▇▇▇▇ 10%sharp
GPQA (subset unspecified) lab46.2% ✎▃······▃·▅·█ beats 2/6hard · avg 43%▇▇▇▇ 10%sharp · at its level
Medmarks-V mean win rate (19 verifiable medical benchmarks, 71 configurations)0.4130 ✎▁▄▃▃▇▄█▃▅▁▁▁ beats 6/30avg 0.423▇▇▇▇ 9.4%sharp
MMLU lab77.8% ✎▄···▄··█··▄█ beats 3/6easy · avg 77%▇▇ 5.2%typical
Medmarks-V weighted mean win rate (61 models, 71 configurations)41.3% ✎▂▃▅▃▆▃█▅█▅▃▆ beats 6/34medium · avg 42%▇▇ 4.6%blunt · at its level
AGIEval English lab81.5% ✎▄····▄·▄▄··█ beats 2/5easy · avg 76%▇▇ 3.8%typical
MMLU-Redux71.3% ✎▁▁▃▁▂▃▄▅▇▅█▄ beats 2/63easy · avg 77%▇ 2.8%at its level
ZebraLogic lab66.5% ✎█▄····▄·▄·▄▄ beats 4/6medium · avg 53%▇ 2.5%at its level
6 more tests
testresultvs the others in its rundifficultycountswhy
GPQA Diamond29.3% ✎▁▁▁▃▂▄▅▅███▄ beats 17/310hard · avg 42%▇ 2.4%sharp · at its level · noisy
PopQA lab23.7% ✎▄··█···▄···█ beats 3/5hard · avg 22%▇ 1.8%blunt
IFBench (aggregation unspecified) lab41.6% ✎█·██·██····█ beats 3/5medium · avg 36%▇ 1.5%at its level
GSM-Plus88.1% ✎▁··▁▁▁·▃▁▄█▁ beats 12/16easy · avg 83%▇ 1.2%blunt
EuroEval HellaSwag (English), test split, 5-shot generative MC, accuracy60.4% ✎▃·▁▁▂▁▂▅█▇▄▁ beats 35/100medium · avg 68%▇ 1.0%at its level · noisy
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy85.7% ✎▁▁▁▁▁▁▁▂▄█▄▃ beats 100/146easy · avg 83%▇ 0.6%noisy

$ 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 Olmo 3 7B Think · needs ~4.5 GB · from huggingface.co/lmstudio-community
    chat in the terminal
    llama cli -hf lmstudio-community/Olmo-3-7B-Think-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf lmstudio-community/Olmo-3-7B-Think-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 Olmo 3 7B Think · needs ~4.5 GB · from huggingface.co/lmstudio-community

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/lmstudio-community/Olmo-3-7B-Think-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 Olmo 3 7B Think · needs ~4.88 GB · from ollama.com
    chat in the terminal
    ollama run olmo-3:7b-think

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 Olmo 3 7B Think · needs ~4.1 GB, Mac only · from huggingface.co/lmstudio-community
    chat in the terminal
    mlx_lm.chat --model lmstudio-community/Olmo-3-7B-Think-MLX-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model lmstudio-community/Olmo-3-7B-Think-MLX-4bit

Weights and documentation: huggingface.co/allenai ↗

$ how the score was worked out · Olmo 3 65%, Medmarks: A Comprehensive Ope… 14%, allenai/Olmo-3-7B-Think model… 10% and 2 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
Olmo 3 · lab-reported14▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 65%CC BY 4.0
Medmarks: A Comprehensive Open-Source LLM Benchmark Suite for Medical Tasks — authors' own runs of 61 models2▇▇▇▇ 14%CC BY 4.0
allenai/Olmo-3-7B-Think model card · lab-reported2▇▇▇ 10%Apache-2.0
JetBrains/Mellum2-12B-A2.5B-Thinking model card5▇▇▇ 9%Apache-2.0
EuroEval leaderboards (results collection), English generative leaderboard2▇ 2%MIT
prior · pull toward the middle▇ 0%
weighted average= 52.9

$ 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