Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Magistral Medium

Mistral

Strongest in Reasoning (#104 of 140, on 1 of its 3 boards), weakest in Safety (#281 of 337). Above par in 0 of 12 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Opus 4.7 and ahead of Grok Build 0.1 and Qwen3 30B A3B Instruct.

§ 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.

Reasoning46#119/178
Reasoning42.7#104/140
Core abilities37.8#198/204
General intelligence32.4#201/204
Human preference47.9#218/342
Human preference47.9#218/342
Safety45.6#281/337
Secure code48.8#170/274
Fairness45.9#177/300
Jailbreak resistance46.8#202/272
Toxicity avoidance47.3#219/272
Harm refusal42.5#273/300

§ 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 Magistral Medium, left for the other.

DBRX Instructwon 7 · lost 1
GPT-3.5 Turbowon 8 · lost 2
Jamba Mini 1.7won 6 · lost 2
Gemma 2B Itwon 6 · lost 2
Smollm2 1.7B Instructwon 6 · lost 2
Zephyr 7B Betawon 6 · lost 2
Gemma 7B Itwon 6 · lost 2
Nous Hermes 2 Mixtral 8x7b Dpowon 6 · lost 2
Llama 3.1 8B Instructwon 6 · lost 2
Command R Pluswon 7 · lost 3
Qwen1.5 14B Chatwon 5 · lost 3
Jamba 1.5 Largewon 5 · lost 3
Qwen3 30B A3B Instructwon 5 · lost 3
Grok Build 0.1won 6 · lost 4
Llama 3.2 1B Instructwon 4 · lost 4
Mistral 7B Instruct V0.2won 4 · lost 4
Llama 3.2 3B Instructwon 4 · lost 4
Granite 3.1 2B Instructwon 4 · lost 4
Gemma 2 2B Itwon 4 · lost 4
Granite 3.1 8B Instructwon 4 · lost 4
Jamba 1.5 Miniwon 4 · lost 4
Gemma 2 9B Itwon 4 · lost 4
Mistral Small 24B Instructwon 4 · lost 4
Nova Prowon 4 · lost 4
Magistral Smallwon 5 · lost 5
Grok 3won 6 · lost 6
Grok 4.3won 6 · lost 6
Mimo V2.6 Flashwon 4 · lost 5
DeepSeek V3won 5 · lost 7
Nemotron 3 Ultra 550B A55Bwon 4 · lost 6
Grok 4.1 Fastwon 4 · lost 6
Ling 3.0 Flash VLwon 3 · lost 5
Phi 3 Mini 128k Instructwon 3 · lost 5
Phi 3 Mini 4k Instructwon 3 · lost 5
Granite 3.0 2B Instructwon 3 · lost 5
Phi 3 Small 8k Instructwon 3 · lost 5
Granite 3.0 8B Instructwon 3 · lost 5
DeepSeek LLM 67B Chatwon 3 · lost 5
Phi 3 Medium 4k Instructwon 3 · lost 5
Llama 3.1 Tulu 3 8Bwon 3 · lost 5
C4ai Aya Expanse 8Bwon 3 · lost 5
Llama 3 8B Instructwon 3 · lost 5
Nova Microwon 3 · lost 5
Nova Litewon 3 · lost 5
C4ai Aya Expanse 32Bwon 3 · lost 5
GPT-4won 3 · lost 5
Kimi K2won 4 · lost 8
DeepSeek R1won 4 · lost 8
Claude 3 Haikuwon 3 · lost 7
Llama 3 70B Instructwon 3 · lost 7
Gemma 3 4B Itwon 3 · lost 7
Gemini 1.5 Prowon 3 · lost 7
Command Awon 3 · lost 8
Mistral Smallwon 3 · lost 8
Jamba Large 1.7won 2 · lost 6
Gemma 4 12Bwon 2 · lost 6
Starling Lm 7B Betawon 2 · lost 6
Phi 4won 2 · lost 6
Qwen2 72B Instructwon 2 · lost 6
Claude 3 Sonnetwon 2 · lost 6
Gemma 2 27B Itwon 2 · lost 6
Qwen2.5 72B Instructwon 2 · lost 6
QwQ 32Bwon 2 · lost 6
Gemini 2.0 Flash Litewon 2 · lost 6
NVIDIA Nemotron 3 Super 120B A12Bwon 2 · lost 6
Gemma 4 26B A4Bwon 2 · lost 6
Llama 4 Maverick Instructwon 3 · lost 9
GPT OSS 120Bwon 3 · lost 9
Gemini 2.0 Flashwon 3 · lost 9
Gemini 2.5 Flashwon 3 · lost 9
Grok 4won 3 · lost 9
DeepSeek V4 Flashwon 3 · lost 9
MiniMax M3won 3 · lost 9
GLM-5.3 Flashwon 3 · lost 9
Mistral Largewon 2 · lost 7
Gemini 1.5 Flashwon 2 · lost 7
Claude 3.5 Haikuwon 2 · lost 7
Kimi K2.7 Codewon 2 · lost 8
Claude 3 Opuswon 2 · lost 8
GPT-4 Turbowon 2 · lost 8
GPT-4owon 2 · lost 8
O1 Miniwon 2 · lost 8
GPT OSS 20Bwon 2 · lost 9
GLM-4.5won 2 · lost 9
GLM-4.6won 2 · lost 9
Gemma 4 31Bwon 2 · lost 9
GPT-4o Miniwon 2 · lost 10
Llama 4 Scout Instructwon 2 · lost 10
GPT-4.1 Nanowon 2 · lost 10
MiniMax M2won 2 · lost 10
GPT-4.1won 2 · lost 10
Qwen3 235B A22B Instructwon 2 · lost 10
Gemini 2.5 Prowon 2 · lost 10
GLM-5.2won 2 · lost 10
Gemma 3 12B Itwon 1 · lost 7
Gemini 2.5 Flash Litewon 1 · lost 7
Qwen3 Vl 235B A22B Instructwon 1 · lost 7
O3 Miniwon 1 · lost 9
Claude 3.5 Sonnetwon 1 · lost 9
Claude 3.7 Sonnetwon 1 · lost 9
Claude Opus 4.6won 1 · lost 9
Mimo V2.5 Prowon 1 · lost 10
GPT-4.1 Miniwon 1 · lost 11
Claude Haiku 4.5won 1 · lost 11
DeepSeek V3.2won 1 · lost 11
Inklingwon 1 · lost 11
Kimi K2.5won 1 · lost 11
GPT-6 Solwon 1 · lost 11
Claude Sonnet 4.5won 1 · lost 11
Kimi K2.6won 1 · lost 11
DeepSeek V4 Prowon 1 · lost 11
Qwen3.7 Maxwon 1 · lost 11
GPT-5.4won 1 · lost 11
Gemini 3.5 Flashwon 1 · lost 11
Gemini 3.6 Flashwon 1 · lost 11
GPT-5.6 Solwon 1 · lost 11
Qwen Maxwon 0 · lost 8
Qwen3.5 35B A3Bwon 0 · lost 8
Qwen3.5 27Bwon 0 · lost 8
MiniMax M2.7won 0 · lost 9
Gemini 3 Prowon 0 · lost 9
Kimi K2 Thinkingwon 0 · lost 10
Gemma 3 27B Itwon 0 · lost 10
O1won 0 · lost 10
Qwen3.5 122B A10Bwon 0 · lost 10
Qwen3.5 397B A17Bwon 0 · lost 10
GPT-5.1won 0 · lost 10
Hy3won 0 · lost 10
GLM-5.1won 0 · lost 10
Claude Sonnet 4.6won 0 · lost 10
Mimo V2.6 Prowon 0 · lost 11
GPT-5 Nanowon 0 · lost 12
GPT-5 Miniwon 0 · lost 12
O4 Miniwon 0 · lost 12
Claude Sonnet 4won 0 · lost 12
Claude Opus 4won 0 · lost 12
O3won 0 · lost 12
GPT-5won 0 · lost 12
GPT-5.2won 0 · lost 12
Grok 4.7won 0 · lost 12
GPT-6 Lunawon 0 · lost 12
Grok 4.6won 0 · lost 12
GPT-5.6 Lunawon 0 · lost 12
Claude Sonnet 5won 0 · lost 12
Grok 4.5won 0 · lost 12
GPT-5.6 Terrawon 0 · lost 12
Qwen3.8 Maxwon 0 · lost 12
Claude Opus 4.8won 0 · lost 12
GPT-5.5won 0 · lost 12
Claude Opus 5won 0 · lost 12
Muse Spark 1.1won 0 · lost 12
Gemini 3.8 Flashwon 0 · lost 12
Claude Opus 4.7won 0 · lost 12
Claude Opus 5.5won 0 · lost 12

§ 3 · Sources

Where the numbers come from

4 publications, 9 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26
LMArena Text 1305
ARC-AGI-2 ↗1 measureread 2026-09-26
ARC-AGI-2 0%
Kagi LLM Benchmark ↗1 measureread 2026-09-26
Kagi LLM Benchmark 16.2%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26
Enkrypt · Jailbreak risk 15.5%Enkrypt · Harmful content risk 66.7%Enkrypt · CBRN risk 16.8%Enkrypt · Toxicity risk 8.2%Enkrypt · Bias risk 84.2%Enkrypt · Insecure code risk 36%

Badge

PublicAI Index badge for Magistral Medium[![PublicAI Index](https://publicai.io/model-index/badge?model=magistral-medium)](https://publicai.io/model-index/m/magistral-medium)