‹ PublicAI Index
The LLM benchmark aggregator.
MiniMax M2.1
MiniMax
Strongest in Academic knowledge (#34 of 123), weakest in Safety (#221 of 329). Above par in 8 of 16 scopes. Among the models it meets almost everywhere, it finishes behind GPT-5.5 and Claude Opus 5 and ahead of GPT-5.4 mini and Gemini 3.1 Flash-Lite.
§ 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.
Knowledge53−9.5#62/1971/3
Academic knowledge57.3−6.6#34/1231/1
Coding50.4−15.8#82/1791/6
Agentic coding50.4−16.5#80/1731/5
Reasoning48.9−16.5#106/1952/5
Mathematics53.1−10.7#70/1521/2
Science51.7−11.8#71/1221/1
Reasoning42.5−26.5#105/1471/3
Professional48.2−14.2#122/1941/2
Medical49−17.5#95/1491/1
Finance48.6−13.8#106/1621/1
Legal46.3−19.6#117/1731/1
Human preference55.1−13.7#151/3541/1
Human preference55.1−13.7#151/3541/1
Safety48.5−12.9#221/3291/3
Factual grounding45.3−25.4#75/1041/1
§ 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 M2.1, left for the other.
§ 3 · Sources
Where the numbers come from
4 publications, 14 figures. Every one links to the page it was read from.
LMArena Text 1384
SimpleBench 34.7%
Vals · CaseLaw 55.84%Vals · LegalBench 75.45%Vals · CorpFin 58.9%Vals · TaxEval 66.35%Vals · MedQA 91.16%Vals · MedCode 34.08%Vals · MedScribe 80.78%Vals · SWE-bench Verified 74.8% (Mini-SWE-agent)Vals · GPQA Diamond 78.54%Vals · MMLU Pro 87.05%Vals · AIME 77.92%
Vectara · Factual consistency 88.2%
Badge
[](https://publicai.io/model-index/m/minimax-m2-1)