[whatmodel.sh](../../README.md) / Mistral Small 4 (119B-A6.5B) # Mistral Small 4 (119B-A6.5B) Mistral · released 2026-03 · 119B params (6.5B active) · apache-2.0 **score 59.4** ±1.5 · #211 of 732 · from 10 benchmark results across 4 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at _{● marks Mistral Small 4 (119B-A6.5B) among every model on that task or test, worst to best.}

task                vs the others            rank  counts  tests
agent tasks         ▁▁▃▆█▄●▅▄▄▂▁     #150 of 313*     19%      1
coding              ██········█●         #1 of 4*    7.3%      1
chat & writing      ▁▁▁▃▆▄●█▆▇▇▃     #264 of 402*    7.5%      1
math                               not ranked yet       —      1
vision & documents  ▁▁▂▂▁▃▄●▆▇█▄      #93 of 153*    7.9%      1

agent tasks · #150 of 313 · 19% of its score

Beats [DeepSeek V3.2 Exp](deepseek-v3-2-exp.md), [o1 (2024-12-17)](o1-2024-12-17.md), [o4-mini](o4-mini.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Large](jamba-1-5-large.md), [GPT-4o (2024-05-13)](gpt-4o-2024-05-13.md) here, though they rank lower overall.

test                                result  vs the others in its run  difficulty           counts  why
ContractBench success rate (33 ob…   42.4%  ▆··▆▄▂▄●▆▆█▄ beats 11/23  very hard · avg 13%     19%  sharp · at its level

coding · #1 of 4 · 7.3% of its score

test                                result  vs the others in its run  difficulty        counts  why
LiveCodeBench (version unspecifie…   63.6%  ▁···▁▁▁█▇●▄▁ beats 28/38  medium · avg 42%    7.3%  at its level · small run

chat & writing · #264 of 402 · 7.5% of its score

Trails [Flan-T5 XXL](flan-t5-xxl.md), [North Micro Vision Instruct (2.4B)](north-micro-vision-instruct.md), [Granite 4.0 H 1B](granite-4-0-h-1b.md) here, though they rank lower overall.

test                                result  vs the others in its run  difficulty        counts  why
Arena-Hard (version/judge unspeci…   58.3%  ▃▃▂▅▃●▅█▅▅▂▃ beats 13/30  medium · avg 47%    7.5%  at its level · small run

math · 0.0% of its score

test               result  vs the others in its run   difficulty        counts  why
AIME 2025 (maker)   83.8%  ▁·▁▁▁▂▅▆█●▅▁ beats 96/133  medium · avg 42%       —  not counted yet: no other model in the …

vision & documents · #93 of 153 · 7.9% of its score

Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GLM-4.6V](glm-4-6v.md), [GPT-5 nano](gpt-5-nano.md) here, though they rank higher overall.

test              result  vs the others in its run  difficulty      counts  why
MMMU Pro (maker)   60.0%  ▂▃▁▃▆▇●▇▇██▅ beats 20/52  hard · avg 41%    7.9%  at its level · small run

other tests (6)

test                                result  vs the others in its run    difficulty           counts  why
GPQA Diamond (maker)                 71.2%  ▁▁▁▁▂▄▅▅●██▄ beats 160/322  hard · avg 42%          16%  sharp · at its level · small run
AraGen-v3 3C3H score (Inception, …   33.9%  ▁▁·▁▂▂▄●▄▅▄▄ beats 31/66    hard · avg 29%          15%  at its level
MMLU-Pro (maker)                     78.0%  ▁▁·▁▁▁▃▂▄●█▁ beats 115/188  medium · avg 59%       9.2%  at its level · small run
Icelandic LLM Leaderboard average…   20.6%  ▁▁▃▂●▄▁▃▄▄▄█ beats 7/38     very hard · avg 14%    7.0%  at its level
IFBench (aggregation unspecified)…   48.0%  ▂▆▂▄▄●▂▄▄▆█▇ beats 36/86    hard · avg 29%         6.6%  at its level · small run
COLLIE (constrained text generati…   62.9%  ▄·▄█▄··▄▄·▄● beats 8/8      medium · avg 42%       4.3%  at its level · small run

## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's 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 is added up by who ran the tests. [Method](../method.md)

source                                    results  share of the score
mistralai/Mistral-Small-4-119B-2603 mod…        7   58% ▇▇▇▇▇▇▇
ContractBench: Can LLM Agents Preserve …        1   19% ▇▇
Inception Arabic/Hindi leaderboards             1   15% ▇▇
Miðeind Icelandic LLM Leaderboard               1    7% ▇
prior: pull toward the middle                        1% ▇
weighted average                                   = 59.4

_{Licences: mistralai/Mistral-Small-4-119B-2603 model card ([Apache-2.0](../../licenses/cards-gap-NOTICE.txt)) · ContractBench: Can LLM Agents Preserve Observation Contracts? ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Miðeind Icelandic LLM Leaderboard ([CC BY 4.0](../../licenses/mideind-icelandic-llm-leaderboard-CC-BY-4.0.txt))}

Not counted (1)

source                  test       score  why
mistralai/Mistral-Sma…  AIME 2025  83.8%  too few models have it yet

## How to run it ### Through an API api (Mistral) · $0.26/M blended · $0.15 in / $0.6 out per 1M tokens · [docs.mistral.ai](https://docs.mistral.ai/models/mistral-small-4-0-26-03) _{[price source](https://github.com/BerriAI/litellm/blob/c19ce71bcdad234ca87a7af10cee45375ce811a8/model_prices_and_context_window.json): USD 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. · [context source](https://github.com/BerriAI/litellm/blob/c19ce71bcdad234ca87a7af10cee45375ce811a8/model_prices_and_context_window.json): Mistral API maximum input tokens from exact catalogue record mistral/mistral-small-2603; not the legacy max\_tokens field.} ### llama.cpp (recommended) 1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md) ```sh # macOS / Linux curl -LsSf https://llama.app/install.sh | sh # Windows irm https://llama.app/install.ps1 | iex ``` 2. Run Mistral Small 4 (119B-A6.5B) (ud-q4_k_m, needs ~71.9 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF/tree/main)) ```sh 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) ```sh llama serve -hf unsloth/Mistral-Small-4-119B-2603-GGUF:UD-Q4_K_M ``` Other versions: ud-q2_k_xl (~36.2 GB) `llama cli -hf unsloth/Mistral-Small-4-119B-2603-GGUF:UD-Q2_K_XL` ### LM Studio 1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download) 2. Run Mistral Small 4 (119B-A6.5B) (ud-q4_k_m, needs ~71.9 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF/tree/main)) ```sh lms get https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF@UD-Q4_K_M -y ``` 3. Then run lms ls to see its name, and lms chat followed by that name Other versions: ud-q2_k_xl (~36.2 GB) `lms get https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF@UD-Q2_K_XL -y` · 4bit (~67.8 GB, Mac only) `lms get https://huggingface.co/mlx-community/Mistral-Small-4-119B-2603-4bit -y` · 8bit (~127.2 GB, Mac only) `lms get https://huggingface.co/mlx-community/Mistral-Small-4-119B-2603-8bit -y` ### Ollama 1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download) ```sh # Linux curl -fsSL https://ollama.com/install.sh | sh ``` 2. Run Mistral Small 4 (119B-A6.5B) (ud-q2_k_xl, needs ~36.2 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Mistral-Small-4-119B-2603-GGUF/tree/main)) ```sh ollama run hf.co/unsloth/Mistral-Small-4-119B-2603-GGUF:UD-Q2_K_XL ``` ### MLX (Mac) 1. Install MLX (Mac). Apple silicon Macs only, text only. ```sh # with uv uv tool install mlx-lm # or pip install mlx-lm ``` 2. Run Mistral Small 4 (119B-A6.5B) (4bit, needs ~67.8 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/Mistral-Small-4-119B-2603-4bit/tree/87b1249207734f029ec9dcefab80eabedfa6b812)) ```sh mlx_lm.chat --model mlx-community/Mistral-Small-4-119B-2603-4bit ``` 3. Or start an OpenAI-compatible server at http://localhost:8080/v1 ```sh mlx_lm.server --model mlx-community/Mistral-Small-4-119B-2603-4bit ``` Other versions: 8bit (~127.2 GB, Mac only) `mlx_lm.chat --model mlx-community/Mistral-Small-4-119B-2603-8bit` Model repository and documentation: [huggingface.co/mistralai](https://huggingface.co/mistralai/Mistral-Small-4-119B-2603) [interactive version](https://whatmodel.sh/model/mistral-small-4) · [JSON](https://whatmodel.sh/api?format=json&all=1)