whatmodel.sh / Mistral Small 4 (119B-A6.5B)

Mistral Small 4 (119B-A6.5B)

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

Mistral Small 4 (119B-A6.5B) is an open-weights language model from Mistral, released March 2026. Its published size is 119B parameters (6.5B active per token) and a 262k-token context window. Licence: apache-2.0.

59.6±1.6#201 of 719 LLMsfrom 10 benchmark results across 4 sources

$ how it compares

The LLMs ranked either side of Mistral Small 4 (119B-A6.5B), out of 828 with a score, then more from Mistral.

#modelscorewhereRAMreleased
240Solar Open 100B (102B-A12B) Upstage60±1local~41GB2025-12
241Qwen3-VL 30B-A3B Thinking Alibaba60±1api $0.75/M +local~21GB2025-09
242Mistral Small 4 (119B-A6.5B) Mistral60±2api $0.26/M +local~40GB2026-03
243Gemini 2.0 Flash-Lite Preview (02-05) Google60±3comparison only2025-02
244QwQ-32B Alibaba59±1local~21GB2025-03
more from Mistral
79Mistral Large 4 (public preview) Mistral71±2api $2.06/M2026-10
174Mistral Medium 3.5 (128B) Mistral65±1api $3.00/M +local~42GB2026-04
180Mistral Medium 3.1 (2508) Mistral64±2comparison only2025-08

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #137 of 300 * · 20% of its score · 1 test

Beats DeepSeek V3.2 Exp, Qwen3 235B-A22B Thinking 2507, o4-mini here, though they rank higher overall. Trails SmolLM3 3B, Jamba 1.5 Large, GPT-4o (2024-05-13) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
ContractBench success rate (33 observation-contract tool tasks)42.4% ✎▆··▆▄▂▄▃▆▆█▄ beats 11/23very hard · avg 14%▇▇▇▇▇▇▇▇ 20%sharp · at its level
coding██········██ #1 of 4 * · 8.4% of its score · 1 test
testresultvs the others in its rundifficultycountswhy
LiveCodeBench (version unspecified) lab63.6% ✎▁···▁▁▃▅▅█▄▅ beats 28/37medium · avg 41%▇▇▇ 8.4%at its level · small run
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #256 of 393 * · 7.9% of its score · 1 test

Trails Flan-T5 XXL, Granite 4.0 H 1B, Granite 4.0 1B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena-Hard (version/judge unspecified) lab58.3% ✎▃▃▂▅▃▆▅█▅▅▂▃ beats 13/30medium · avg 50%▇▇▇ 7.9%at its level · small run
mathnot ranked on this task yet · 1 test
testresultvs the others in its rundifficultycountswhy
AIME 2025 lab83.8% ✎▁·▁▁▁▂▅▆█▇▅▁ beats 96/133medium · avg 42%—not counted yet: no other model in the same run to compare it with
vision & documents▁▁▃▂▁▃▄▃▆▇█▄ #87 of 147 * · 8.0% of its score · 1 test

Beats DeepSeek V4 Flash Vision (Exp), GLM-4.6V, GPT-5 nano here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
MMMU Pro lab60.0% ✎▂▃▁▃▆▇▃▇▇██▅ beats 20/52hard · avg 42%▇▇▇ 8.0%at its level · small run
other tests6 not in a task · 55% of its score
testresultvs the others in its rundifficultycountswhy
GPQA Diamond lab71.2% ✎▁▁▁▁▂▄▅▅███▄ beats 157/310hard · avg 43%▇▇▇▇▇▇ 15%sharp · at its level · small run
AraGen-v3 3C3H score (Inception, Claude Sonnet 4 judge)33.9% ✎▁▁·▁▂▂▄█▄▅▄▄ beats 31/66hard · avg 31%▇▇▇▇▇ 12%at its level
MMLU-Pro lab78.0% ✎▁▁·▁▁▁▃▂▄▅█▁ beats 115/188medium · avg 60%▇▇▇▇ 9.5%at its level · small run
IFBench (aggregation unspecified) lab48.0% ✎▃▇▂▄▅▅▂▄▄▆█▇ beats 36/80hard · avg 31%▇▇▇ 6.9%at its level · small run
Icelandic LLM Leaderboard average (Miðeind, Sept 2026 suite)20.6% ✎▁▁▃▂▃▄▁▃▄▄▄█ beats 7/38very hard · avg 14%▇▇ 6.2%at its level
COLLIE (constrained text generation) lab62.9% ✎▄·▄█▄··▄▄·▄▄ beats 8/8medium · avg 41%▇▇ 4.6%at its level · small run

$ where to run it

Through an API

api · $0.26/M · $0.15 in / $0.6 out per 1M tokens · 262k context · docs.mistral.ai ↗ · price sourceUSD per 1M text tokens. Mistral standard synchronous API; model mistral-small-2603. Uncached input; output includes generated reasoning where billed. Batch, cache storage, tools and other modalities excluded.

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 Mistral Small 4 (119B-A6.5B) · needs ~71.9 GB · from huggingface.co/unsloth
    chat in the terminal
    llama cli -hf unsloth/Mistral-Small-4-119B-2603-GGUF:UD-Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf unsloth/Mistral-Small-4-119B-2603-GGUF:UD-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 Mistral Small 4 (119B-A6.5B) · needs ~71.9 GB · from huggingface.co/unsloth

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF@UD-Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Mistral Small 4 (119B-A6.5B) · needs ~36.2 GB · from huggingface.co/unsloth
    chat in the terminal
    ollama run hf.co/unsloth/Mistral-Small-4-119B-2603-GGUF:UD-Q2_K_XL

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 Mistral Small 4 (119B-A6.5B) · needs ~67.8 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/Mistral-Small-4-119B-2603-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/Mistral-Small-4-119B-2603-4bit

Weights and documentation: huggingface.co/mistralai ↗

$ how the score was worked out · mistralai/Mistral-Small-4-119… 61%, ContractBench: Can LLM Agents… 20%, Inception Arabic/Hindi leader… 12% 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
mistralai/Mistral-Small-4-119B-2603 model card · lab-reported7▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 61%Apache-2.0
ContractBench: Can LLM Agents Preserve Observation Contracts? (arXiv:2605.17281v2)1▇▇▇▇▇▇ 20%CC BY 4.0
Inception Arabic/Hindi leaderboards (AraGen-v3, HindiGen-v1, Arabic IFEval)1▇▇▇▇ 12%Apache-2.0
Miðeind Icelandic LLM Leaderboard1▇▇ 6%CC BY 4.0
prior · pull toward the middle▇ 1%
weighted average= 59.6

not counted (1)

mistralai/Mistral-Small-4-119B-2603 model cardAIME 2025 [correct]83.8%fewer than three models have this test yet, or this source ran it for this model alone

$ 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