Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Mistral Small

Mistral

Strongest in Factual grounding (#12 of 101), weakest in Harm refusal (#236 of 300). Above par in 5 of 21 scopes. Among the models it meets almost everywhere, it finishes behind Claude Sonnet 5 and O3 and ahead of Llama 4 Scout Instruct and Claude 3.5 Haiku.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Agents49.7#123/268
Tool use49.5#37/81
Knowledge32.5#132/138
Academic knowledge29#116/123
Core abilities44.9#145/204
General intelligence42.7#150/204
Professional40.1#155/168
Medical35.9#129/140
Finance35#142/151
Human preference52.9#159/342
Human preference52.9#159/342
Reasoning36.2#167/178
Science32.4#107/122
Mathematics33.9#133/140
Safety48.7#219/337
Factual grounding62.3#12/101
Jailbreak resistance52.7#147/272
Toxicity avoidance51.1#189/272
Fairness44.9#200/300
Secure code40.7#216/274
Harm refusal45.6#236/300

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Mistral Small, left for the other.

Jamba 1.6 Miniwon 13 · lost 1
Command R Pluswon 12 · lost 3
Jamba 1.6 Largewon 10 · lost 4
Gemma 3 4B Itwon 8 · lost 5
Claude 3.5 Haikuwon 9 · lost 7
Llama 4 Scout Instructwon 11 · lost 10
GPT-4.1 Nanowon 10 · lost 10
Command Awon 10 · lost 11
Mistral Medium 3.5won 6 · lost 7
GPT-4o Miniwon 8 · lost 11
Gemini 1.5 Flashwon 6 · lost 9
Gemma 3 27B Itwon 5 · lost 8
Ling 3.0 Flashwon 5 · lost 9
Mistral Largewon 6 · lost 13
DeepSeek R1won 6 · lost 13
Grok 4.1 Fastwon 6 · lost 13
Grok 4.3won 6 · lost 13
GPT OSS 120Bwon 6 · lost 14
DeepSeek V3won 6 · lost 14
Gemma 4 31Bwon 5 · lost 13
Mistral Large 3won 4 · lost 11
GPT OSS 20Bwon 5 · lost 14
MiniMax M3won 5 · lost 15
Kimi K2.6won 5 · lost 15
GLM-4.6won 5 · lost 16
GPT-4owon 4 · lost 13
Gemini 3.5 Flash Litewon 3 · lost 10
GPT-5.4 Miniwon 3 · lost 11
Grok 3won 4 · lost 15
GLM-5.2won 4 · lost 16
Gemini 1.5 Prowon 3 · lost 13
GPT-6 Lunawon 3 · lost 13
GPT-6 Solwon 3 · lost 14
Gemini 2.0 Flashwon 3 · lost 15
GLM-4.5won 3 · lost 15
DeepSeek V4 Flashwon 3 · lost 16
Qwen3.7 Maxwon 3 · lost 16
Gemini 3.5 Flashwon 3 · lost 16
Gemini 3.6 Flashwon 3 · lost 16
Qwen3.6 Pluswon 2 · lost 11
Grok 4won 3 · lost 17
Gemini 2.5 Prowon 3 · lost 17
Grok 4 Fastwon 2 · lost 12
GPT-5 Nanowon 3 · lost 18
Gemini 2.5 Flashwon 3 · lost 18
Kimi K2won 3 · lost 18
DeepSeek V3.2won 3 · lost 18
GPT-5.4 Nanowon 2 · lost 13
Gemini 3.1 Flash Litewon 2 · lost 13
Gemini 3.1 Prowon 2 · lost 13
Claude 3.7 Sonnetwon 2 · lost 15
Gemini 2.5 Flash Litewon 2 · lost 16
Claude Sonnet 4won 2 · lost 18
Kimi K2.5won 2 · lost 18
DeepSeek V4 Prowon 2 · lost 18
GPT-5.4won 2 · lost 18
GPT-5.5won 2 · lost 18
GPT-5.6 Solwon 2 · lost 18
O4 Miniwon 2 · lost 19
Qwen3 235B A22B Instructwon 2 · lost 19
GLM-5won 1 · lost 13
Kimi K2 Thinkingwon 1 · lost 15
Claude Opus 4.5won 1 · lost 15
O3 Miniwon 1 · lost 16
GLM-5.1won 1 · lost 16
Claude Opus 4.6won 1 · lost 16
GPT-5.1won 1 · lost 17
Claude Sonnet 4.6won 1 · lost 17
Gemini 3 Prowon 1 · lost 17
Claude Opus 4won 1 · lost 18
Grok 4.5won 1 · lost 18
Mimo V2.5 Prowon 1 · lost 18
GLM-5.3 Flashwon 1 · lost 18
Claude Opus 5won 1 · lost 18
Gemini 3.8 Flashwon 1 · lost 18
GPT-4.1 Miniwon 1 · lost 19
GPT-5won 1 · lost 19
Inklingwon 1 · lost 19
GPT-5.6 Lunawon 1 · lost 19
GPT-5.6 Terrawon 1 · lost 19
Claude Opus 4.7won 1 · lost 19
GPT-5 Miniwon 1 · lost 20
Claude Haiku 4.5won 1 · lost 20
GPT-4.1won 1 · lost 20
GPT-5.2won 1 · lost 20
Claude Sonnet 4.5won 1 · lost 20
GLM-5.3won 0 · lost 13
Gemini 3.7 Flashwon 0 · lost 13
Kimi K3won 0 · lost 13
Claude Fable 5.1won 0 · lost 13
Claude Fable 5won 0 · lost 13
Qwen3.8 27Bwon 0 · lost 14
Mimo V2.6 Prowon 0 · lost 15
Claude 3.5 Sonnetwon 0 · lost 16
O1won 0 · lost 16
Grok 4.7won 0 · lost 16
Claude Opus 5.5won 0 · lost 16
MiniMax M2.7won 0 · lost 17
Grok 4.6won 0 · lost 19
Qwen3.8 Maxwon 0 · lost 19
Claude Opus 4.8won 0 · lost 19
Muse Spark 1.1won 0 · lost 19
O3won 0 · lost 20
Claude Sonnet 5won 0 · lost 20

§ 3 · Sources

Where the numbers come from

6 publications, 15 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26
LMArena Text 1357
BFCL ↗1 measureread 2026-09-26
BFCL v4 37.15%
Kagi LLM Benchmark ↗1 measureread 2026-09-26
Kagi LLM Benchmark 37.8%
Vals.ai ↗5 measuresread 2026-09-26
Vals · TaxEval 49.14%Vals · MedQA 56.98%Vals · GPQA Diamond 50.76%Vals · MMLU Pro 64.44%Vals · AIME 5.63%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26
Enkrypt · Jailbreak risk 10.5%Enkrypt · Harmful content risk 60.6%Enkrypt · CBRN risk 11.8%Enkrypt · Toxicity risk 5.5%Enkrypt · Bias risk 85.8%Enkrypt · Insecure code risk 52.4%
Vectara · Factual consistency 94.9%

Badge

PublicAI Index badge for Mistral Small[![PublicAI Index](https://publicai.io/model-index/badge?model=mistral-small)](https://publicai.io/model-index/m/mistral-small)