The LLM benchmark aggregator.

MiniMax M2.7

MiniMax · 229B · open weights

Strongest in Safety (#3 of 337, on 1 of its 3 boards), weakest in Coding (#126 of 165). Above par in 13 of 21 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Claude Fable 5.1 and ahead of Kimi K2.5 and Gemini 2.5 Pro.

§ 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.

Safety60.5−1#3/3371/3

Secure code66.1−0.4#6/2741/1

Fairness70leads#11/3001/2

Jailbreak resistance62.5−2.4#18/2721/1

Harm refusal56.3−5.3#43/3001/2

Toxicity avoidance56.4−2.6#61/2721/1

Reasoning55.7−11.2#34/1781/4

Mathematics56.6−9.1#35/1401/2

Science57.3−6.2#42/1221/1

Knowledge49.2−14.3#90/1381/2

Academic knowledge49−14.9#83/1231/1

Agents51.1−17#106/2682/5

Knowledge work51.4−22.2#72/1782/2

Human preference58.5−9#107/3421/1

Professional45.7−15.9#124/1681/1

Legal46.4−20.7#96/1511/1

Medical46.1−20.7#105/1401/1

Finance43.6−19.3#119/1511/1

Coding43.6−25.5#126/1651/5

Agentic coding42.9−25.8#121/1571/4

§ 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 MiniMax M2.7, left for the other.

Jamba 1.6 Miniwon 15 · lost 0

Jamba 1.6 Largewon 15 · lost 0

Command R Pluswon 15 · lost 0

Gemini 1.5 Flashwon 16 · lost 0

Mistral Smallwon 17 · lost 0

Llama 4 Scout Instructwon 19 · lost 0

GPT-4o Miniwon 20 · lost 0

GPT-4.1 Nanowon 21 · lost 0

GPT OSS 20Bwon 17 · lost 2

Gemini 1.5 Prowon 15 · lost 2

Mistral Medium 3.5won 13 · lost 2

Llama 3.3 Instruct Turbo (70B)won 12 · lost 2

Llama 4 Maverick Instructwon 18 · lost 3

Ling 3.0 Flashwon 15 · lost 3

Mistral Largewon 14 · lost 3

GPT-5.4 Nanowon 13 · lost 3

Gemini 2.5 Flash Litewon 12 · lost 3

Command Awon 16 · lost 4

Gemini 2.0 Flashwon 15 · lost 4

GPT-5 Nanowon 14 · lost 4

Grok 3won 13 · lost 4

GPT OSS 120Bwon 16 · lost 5

O3 Miniwon 16 · lost 5

Mistral Large 3won 12 · lost 4

GPT-4.1 Miniwon 15 · lost 6

DeepSeek V3won 15 · lost 6

Claude Haiku 4.5won 15 · lost 6

Claude 3.5 Haikuwon 14 · lost 6

Gemma 4 31Bwon 12 · lost 6

GLM-4.5won 11 · lost 6

Grok 4 Fastwon 9 · lost 5

GPT-4owon 12 · lost 7

GPT-5 Miniwon 13 · lost 8

Claude Sonnet 4won 13 · lost 8

Grok 4.3won 13 · lost 8

GPT-4 Turbowon 8 · lost 5

O1 Miniwon 8 · lost 5

Gemini 3.1 Flash Litewon 9 · lost 7

Gemini 2.5 Flashwon 11 · lost 9

GPT-4.1won 11 · lost 9

DeepSeek R1won 11 · lost 9

Grok 4.1 Fastwon 11 · lost 9

MiniMax M2.5won 7 · lost 6

Mimo V2.6 Flashwon 8 · lost 7

Kimi K2 Thinkingwon 9 · lost 8

GLM-4.6won 11 · lost 10

DeepSeek V3.2won 11 · lost 10

Gemini 2.5 Prowon 11 · lost 10

Kimi K2.5won 11 · lost 10

Claude 3.5 Sonnetwon 9 · lost 9

O4 Miniwon 10 · lost 10

Grok 4won 10 · lost 10

Kimi K2won 10 · lost 10

O1won 9 · lost 10

Qwen3 235B A22B Instructwon 9 · lost 10

Mimo V2.6 Prowon 8 · lost 9

MiniMax M2.1won 6 · lost 7

Qwen3 Maxwon 6 · lost 7

Claude 3.7 Sonnetwon 9 · lost 11

GPT-6 Lunawon 8 · lost 10

Qwen3.5 Flashwon 6 · lost 8

MiniMax M3won 9 · lost 12

GLM-5.1won 9 · lost 12

Mimo V2.5 Prowon 9 · lost 12

GLM-5.3 Flashwon 9 · lost 12

GLM-5.2won 9 · lost 12

Inkling Smallwon 6 · lost 9

Mimo V2.5won 6 · lost 9

GPT-5.4 Miniwon 6 · lost 9

Gemini 3.5 Flash Litewon 6 · lost 9

DeepSeek V4 Flashwon 8 · lost 13

GPT-5.6 Lunawon 8 · lost 13

Claude Sonnet 4.5won 8 · lost 13

Kimi K2.6won 8 · lost 13

DeepSeek V4 Prowon 8 · lost 13

Qwen3.7 Maxwon 8 · lost 13

Gemini 3.5 Flashwon 8 · lost 13

Gemini 3.6 Flashwon 8 · lost 13

Claude Opus 4won 7 · lost 12

O3won 7 · lost 13

Grok 4.7won 6 · lost 12

GPT-6 Solwon 6 · lost 12

Inklingwon 7 · lost 14

Grok 4.5won 7 · lost 14

Qwen3.8 Maxwon 7 · lost 14

Gemini 3.8 Flashwon 7 · lost 14

Claude Sonnet 4.6won 6 · lost 13

GPT-5won 6 · lost 15

GPT-5.1won 6 · lost 15

GPT-5.6 Terrawon 6 · lost 15

GPT-5.4won 6 · lost 15

GPT-5.5won 6 · lost 15

GPT-5.6 Solwon 6 · lost 15

Muse Spark 1.1won 6 · lost 15

GLM-4.7won 4 · lost 11

Qwen3.6 Pluswon 4 · lost 11

Claude Opus 4.5won 4 · lost 11

Gemini 3.1 Prowon 4 · lost 12

GPT-5.2won 5 · lost 15

Grok 4.6won 5 · lost 16

Claude Opus 4.6won 4 · lost 13

GLM-5won 3 · lost 11

DeepSeek V4.1 Flashwon 3 · lost 11

Qwen3.8 27Bwon 3 · lost 12

Claude Opus 4.8won 4 · lost 17

Claude Opus 5won 4 · lost 17

Gemini 3 Prowon 3 · lost 14

Claude Opus 4.7won 3 · lost 18

GLM-5.3won 2 · lost 13

Muse Sparkwon 2 · lost 13

Claude Opus 5.5won 2 · lost 16

GPT-6 Astrawon 1 · lost 12

Gemini 3 Flashwon 1 · lost 13

Muse Spark 1.2won 1 · lost 13

Gemini 3.7 Flashwon 1 · lost 14

Kimi K3won 1 · lost 14

Claude Sonnet 5won 1 · lost 20

Claude Fable 5.1won 0 · lost 15

Claude Fable 5won 0 · lost 15

§ 3 · Sources

Where the numbers come from

5 publications, 24 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26

LMArena Text 1415

GDPval-AA ↗1 measureread 2026-09-26

GDPval-AA 999

AA-Briefcase ↗1 measureread 2026-09-26

AA-Briefcase 716

Vals.ai ↗15 measuresread 2026-09-26

Vals · Legal Research Bench 10.58%Vals · CaseLaw 60.88%Vals · LegalBench 83.98%Vals · Harvey Legal Agent Benchmark 0%Vals · Finance Agent 27.89%Vals · CorpFin 61.19%Vals · TaxEval 66.56%Vals · MedCode 34.44%Vals · MedScribe 79.87%Vals · SWE-bench Verified 73.8% (Mini-SWE-agent)Vals · Vibe Code Bench 11.93% (OpenHands)Vals · Code Migration 8.69%Vals · GPQA Diamond 86.62%Vals · MMLU Pro 80.43%Vals · AIME 91.04%

Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26

Enkrypt · Jailbreak risk 2.2%Enkrypt · Harmful content risk 2.2%Enkrypt · CBRN risk 15%Enkrypt · Toxicity risk 1.8%Enkrypt · Bias risk 45.5%Enkrypt · Insecure code risk 0.9%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=minimax-m2-7)](https://publicai.io/model-index/m/minimax-m2-7)