MiniMax M3vsMiniMax M2.5

Across 3 shared benchmarks, MiniMax M3 leads overall: MiniMax M3 wins 3, MiniMax M2.5 wins 0, with 0 ties and an average score difference of +6.89.

MiniMaxAI
MiniMax M3

MiniMaxAI · 2026-06-01 · Multimodal model

MiniMaxAI
MiniMax M2.5

MiniMaxAI · 2026-02-12 · Reasoning model

MiniMax M33 wins(100%)(0%)0 winsMiniMax M2.5

Benchmark scores

Grouped by capability, sorted by largest gap within each. 3 shared benchmarks.

AI Agent - Information Search

MiniMax M3 1/1
BenchmarkMiniMax M3MiniMax M2.5Diff
BrowseComp83.508 / 45Thinking (With Tools + Internet)76.3018 / 45+7.20

Coding and Software Engineer

MiniMax M3 1/1
BenchmarkMiniMax M3MiniMax M2.5Diff
SWE-Bench Pro - Public597 / 44Thinking (With Tools)55.4019 / 44+3.60

General Knowledge

MiniMax M3 1/1
BenchmarkMiniMax M3MiniMax M2.5Diff
LiveBench70.0240 / 115Deep Thinking (No Tools)60.1468 / 115Deep Thinking (No Tools)+9.88

Specs

FieldMiniMax M3MiniMax M2.5
PublisherMiniMaxAIMiniMaxAI
Release date2026-06-012026-02-12
Model typeMultimodal modelReasoning model
ArchitectureMoEMoE
Parameters428B229B
Context length1M128K
Max output512KNot available

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemMiniMax M3MiniMax M2.5
Text input¥2.1 / 1M tokens$0.3 / 1M tokens
Text output¥8.4 / 1M tokens$2.4 / 1M tokens
Cache read¥0.42 / 1M tokensNot public

Summary

  • MiniMax M3leads in:AI Agent - Information Search (1/1), Coding and Software Engineer (1/1), General Knowledge (1/1)

On average across the 3 shared benchmarks, MiniMax M3 scores 6.89 higher.

Largest single-benchmark gap: LiveBench — MiniMax M3 70.02 vs MiniMax M2.5 60.14 (+9.88).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.