Codesota · Models · MiniMax-M2.5MiniMaxAI11 results · 5 benchmarks
Model card

MiniMax-M2.5.

MiniMaxAIopen-source
§ 02 · Benchmarks

Every benchmark MiniMax-M2.5 has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01SWE-Bench VerifiedComputer Code · Code Generationaccuracy80.2%#2/22source ↗
02BrowseCompNatural Language Processing · Question Answeringaccuracy76.3%#3/16source ↗
03GPQA DiamondReasoning · Multi-step Reasoningaccuracy85.2%#21/74source ↗
04HLEReasoning · Multi-step Reasoningaccuracy19.4%#38/74source ↗
05PLCCNatural Language Processing · Polish Cultural Competencygrammar71.0%#47/165source ↗
06PLCCNatural Language Processing · Polish Cultural Competencyvocabulary52.0%#89/165source ↗
07PLCCNatural Language Processing · Polish Cultural Competencyaverage59.7%#90/165source ↗
08PLCCNatural Language Processing · Polish Cultural Competencyculture-and-tradition59.0%#91/165source ↗
09PLCCNatural Language Processing · Polish Cultural Competencygeography68.0%#95/165source ↗
10PLCCNatural Language Processing · Polish Cultural Competencyhistory69.0%#97/165source ↗
11PLCCNatural Language Processing · Polish Cultural Competencyart-and-entertainment39.0%#117/165source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area

Where MiniMax-M2.5 actually performs.

Computer Code
1
benchmark
avg rank #2.0
Reasoning
2
benchmarks
avg rank #29.5
Natural Language Processing
2
benchmarks
avg rank #78.6
§ 04 · Papers

1 paper with results for MiniMax-M2.5.

  1. 2026-02-12· 4 results

    MiniMax-M2.5

§ 05 · Related models

Other MiniMaxAI models scored on Codesota.

MiniMax-M2.7
0 results
§ 06 · Sources & freshness

Where these numbers come from.

sdadas/PLCC
7
results
pwc-dump
4
results
7 of 11 rows marked verified.