whatmodel.sh / Magistral Small 1.2

Magistral Small 1.2

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

Magistral Small 1.2 is an open-weights language model from Mistral, released September 2025. Its published size is 24B parameters and a 131k-token context window. Licence: apache-2.0.

57.8±1.1#219 of 719 LLMsfrom 14 benchmark results across 6 sources

$ how it compares

The LLMs ranked either side of Magistral Small 1.2, out of 828 with a score, then more from Mistral.

#modelscorewhereRAMreleased
261Phi-4-reasoning 14B Microsoft58±1api $0.22/M +local~10GB2025-04
262Olmo 3 32B Think Ai258±1local~21GB2025-11
263Magistral Small 1.2 Mistral58±1api $0.75/M +local~16GB2025-09
264Nemotron-Cascade-8B-Thinking NVIDIA58±2local~6.3GB2025-12
265Qwen3-VL 30B-A3B Instruct Alibaba58±1api $0.35/M +local~21GB2025-09
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

coding██········██ #2 of 4 * · 9.3% of its score · 2 tests
testresultvs the others in its rundifficultycountswhy
LiveCodeBench v5 (window unspecified) lab70.9% ✎▄█▄······▄·▄ beats 4/5medium · avg 50%▇▇ 5.5%at its level
LiveCodeBench (version unspecified) lab60.7% ✎▁···▁▁▃▅▅█▄▅ beats 25/37medium · avg 41%▇▇ 3.9%at its level · small run
math▁▂▃▅▆▅▄█▄▁▁▁ #179 of 398 · 37% of its score · 3 tests

Beats MiniMax-M2.1, GPT-4.1, DeepSeek V3 0324 here, though they rank higher overall. Trails Phi-4-mini-reasoning 3.8B, OpenThinker3 7B, Olmo 3 7B Think here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
AIME 2025 lab77.3% ✎█▃·····▃···▃ beats 4/5medium · avg 49%▇▇▇▇▇▇ 15%sharp · at its level
AIME 2024 lab86.1% ✎█▃·····▃···▃ beats 4/5medium · avg 60%▇▇▇▇▇ 12%sharp
OTIS mock AIME 2024–202528.1% ✎▃▃▄▃▃▃▇▇█▃▂▅ beats 62/193very hard · avg 8.5%▇▇▇▇ 9.4%sharp · at its level · noisy
vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #97 of 147 * · 8.4% of its score · 2 tests

Beats DeepSeek V4 Flash Vision (Exp) here, though they rank higher overall. Trails Llama 4 Maverick (402B-A17B), GPT-4o (2024-11-20), Qwen3-VL 8B Thinking here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
CharXiv (reasoning questions)52.9% ✎▂▅▃▅▅▆▆▅▆█▆▂ beats 33/100medium · avg 38%▇▇ 4.6%at its level · small run
CharXiv descriptive questions82.3% ✎▂▄▄▅▇▄█▇▇▂▁▂ beats 28/58easy · avg 72%▇▇ 3.8%small run
other tests6 not in a task · 45% of its score
testresultvs the others in its rundifficultycountswhy
GPQA Diamond +147.6% ✎▁▁▁▁▂▄▆▄▅▇█▅ beats 60/208hard · avg 40%▇▇▇▇▇▇▇▇▇▇▇▇▇ 34%sharp · at its level · several runs
AraGen-v3 3C3H score (Inception, Claude Sonnet 4 judge)32.3% ✎▁▁·▁▂▂▄█▄▅▄▄ beats 29/66hard · avg 31%▇▇ 5.7%at its level
COLLIE (constrained text generation) lab60.3% ✎▄·▄█▄··▄▄·▄▄ beats 7/8medium · avg 41%▇ 2.1%at its level · small run
EuroEval HellaSwag (English), test split, 5-shot generative MC, accuracy89.4% ✎▃·▁▁▂▁▂▅█▇▄▁ beats 90/100medium · avg 68%▇ 1.4%at its level · noisy
Chess Puzzles (Epoch)3.0% ✎█▁▁▁▁▂▁▄▂▂▁▁ beats 38/135very hard · avg 5.9%▇ 1.2%sharp · too hard for it · noisy
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy95.5% ✎▁▁▁▁▁▁▁▂▄█▄▃ beats 142/146easy · avg 83%▇ 0.8%noisy

$ where to run it

Through an API

api · Amazon Bedrock · $0.75/M · $0.5 in / $1.5 out per 1M tokens · 131k context · docs.aws.amazon.com ↗ · price sourceStandard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ.

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 Magistral Small 1.2 · needs ~14.9 GB · from huggingface.co/mistralai
    chat in the terminal
    llama cli -hf mistralai/Magistral-Small-2509-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf mistralai/Magistral-Small-2509-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 Magistral Small 1.2 · needs ~14.9 GB · from huggingface.co/mistralai

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/mistralai/Magistral-Small-2509-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 Magistral Small 1.2 · needs ~14.9 GB · from huggingface.co/mistralai
    chat in the terminal
    ollama run hf.co/mistralai/Magistral-Small-2509-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 Magistral Small 1.2 · needs ~14.1 GB, Mac only · from huggingface.co/lmstudio-community
    chat in the terminal
    mlx_lm.chat --model lmstudio-community/Magistral-Small-2509-MLX-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model lmstudio-community/Magistral-Small-2509-MLX-4bit

Weights and documentation: huggingface.co/mistralai ↗

$ how the score was worked out · mistralai/Magistral-Small-250… 48%, Epoch AI — internal evaluatio… 29%, Apriel-1.5-15b-Thinker (arXiv… 8% and 3 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/Magistral-Small-2509 model card · lab-reported4▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 48%Apache-2.0
Epoch AI — internal evaluations3▇▇▇▇▇▇▇▇▇ 29%CC BY 4.0
Apriel-1.5-15b-Thinker (arXiv:2510.01141v1), Table 4 multimodal benchmarks2▇▇▇ 8%CC-BY-4.0
mistralai/Mistral-Small-4-119B-2603 model card · lab-reported2▇▇ 6%Apache-2.0
Inception Arabic/Hindi leaderboards (AraGen-v3, HindiGen-v1, Arabic IFEval)1▇▇ 6%Apache-2.0
EuroEval leaderboards (results collection), English generative leaderboard2▇ 2%MIT
prior · pull toward the middle▇ 1%
weighted average= 57.8

$ 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