Compare / MiniMax M3 vs Kimi K2.5

MiniMax M3 vs Kimi K2.5

15 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.

MiniMax M3
MiniMax · profile
Kimi K2.5
Moonshot AI · profile
Results compared
15
Shared sources
3
BenchmarkMetricMiniMax M3Kimi K2.5
Arena (formerly LMArena)reportedOverallArena rating 1,4331,436
Arena (formerly LMArena)reportedOverallArena rating 1,4341,445
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4501,441
Arena (formerly LMArena)reportedCreative writingArena rating 1,4061,423
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4711,483
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4351,439
Arena (formerly LMArena)reportedOverallArena rating 1,4401,450
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4221,428
OpenHands IndexreportedAverage% resolved 57.2%49.2%
OpenHands IndexreportedFront end (SWE-Bench Multimodal)% resolved 36.8%32.8%
OpenHands IndexreportedGreenfield (Commit0)% resolved 25.0%18.8%
OpenHands IndexreportedInformation gathering (GAIA)% resolved 66.7%63.6%
OpenHands IndexreportedIssue resolution (SWE-Bench)% resolved 76.4%68.8%
OpenHands IndexreportedTesting (SWT-Bench)% resolved 81.1%61.9%
Vending-Bench 2reportedMoney after a yearUS dollars $2,157.77$1,198.46

Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.