MMLU-Pro ↗87.49%
MiniMax-M2.1 · MMLU-Pro · AA implementation ↗
1 more reported result
87.05 % · MiniMax-M2.1 · MMLU Pro · Accuracy · Vals AI · MMLU Pro ↗
MiniMax / MODEL EVIDENCE
MiniMax M2.1
18 benchmarks · 21 observations · 20 sources · released 2025-12-23
Scores at a glance
18 benchmarks · large dot: MiniMax M2.1; small dots: closest models where reported
LiveCodeBench ↗80.95%
MiniMax-M2.1 · LiveCodeBench · AA implementation ↗
1 more reported result
81.76 % · MiniMax-M2.1 · LiveCodeBench · Accuracy · Vals AI · LiveCodeBench ↗
Terminal-Bench 2.0 ↗36.63%
MiniMax-M2.1 · Terminal Bench · Accuracy mean · % ↗
1 more reported result
29.21% · MiniMax-M2.1 · Terminal Bench · Accuracy mean · % · Epoch AI · Terminal Bench ↗
Text Arena (Coding) ↗1387.48 Elo
MiniMax-M2.1 · WebDev Arena · Elo ↗
LegalBench ↗75.45 %
MiniMax-M2.1 · LegalBench · Accuracy ↗
CaseLaw v2 ↗55.84 %
MiniMax-M2.1 · CaseLaw · Accuracy ↗
CorpFin v2 ↗58.9 %
MiniMax-M2.1 · CorpFin · Accuracy ↗
MortgageTax & TaxEval v2 ↗66.35 %
MiniMax-M2.1 · TaxEval · Accuracy ↗
MedCode ↗34.08 %
MiniMax-M2.1 · MedCode · Accuracy ↗
MedScribe ↗80.78 %
MiniMax-M2.1 · MedScribe · Accuracy ↗
MedQA (USMLE) ↗91.16 %
MiniMax-M2.1 · MedQA · Accuracy ↗
AIME 2024–2025 · Vals ↗77.92 %
MiniMax-M2.1 · AIME · Accuracy ↗
MATH-500 ↗89 %
MiniMax-M2.1 · MATH 500 · Accuracy ↗
MGSM (Multilingual GSM) ↗87.85 %
MiniMax-M2.1 · MGSM · Accuracy ↗
GPQA Diamond ↗78.54 %
MiniMax-M2.1 · GPQA Diamond · Accuracy ↗
SWE-bench Verified ↗74.8 %
MiniMax-M2.1 (Mini-SWE-agent) · SWE-bench Verified · Score ↗
NYU CTF Bench ↗12.3%
MiniMax M2.1 · NYU CTF Bench · Cotool 81-challenge subset · accuracy ↗
ALE-Bench ↗623.83 Performance points
minimax-m2.1 · Average contest performance ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
9 of 14 shared benchmark comparisons
9 of 14 shared benchmark comparisons
9 of 14 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- Arena leaderboard — WebDev evaluation ↗
- LiveCodeBench Benchmark | Artificial Analysis ↗
- MMLU-Pro Benchmark | Artificial Analysis ↗
- Ale Bench — methodology and published results ↗
- NYU CTF Bench — Cotool evaluation, February 2026 ↗
- Terminal-Bench 2.0 leaderboard ↗
- AIME — evaluation methodology and results ↗
- CaseLaw — evaluation methodology and results ↗
- CorpFin — evaluation methodology and results ↗
- GPQA Diamond — evaluation methodology and results ↗
- LiveCodeBench — evaluation methodology and results ↗
- LegalBench — evaluation methodology and results ↗
- MATH 500 — evaluation methodology and results ↗
- MedCode — evaluation methodology and results ↗
- MedQA — evaluation methodology and results ↗
- MedScribe — evaluation methodology and results ↗
- MGSM — evaluation methodology and results ↗
- MMLU Pro — evaluation methodology and results ↗
- SWE-bench Verified — evaluation methodology and results ↗
- TaxEval — evaluation methodology and results ↗
Names used in source tables
MiniMax M2.1 · MiniMax-M2.1 · MiniMax-M2.1 (Mini-SWE-agent) · minimax-m2.1