MiniMax-M2

MiniMax-M2 ranks first on ArtifactsBench.

MiniMax Released 27 Oct 2025 Family MiniMax-M2 The lab's model card

Benchmarks with a score

Ranked first, out of five or more models

In the top three

MiniMax Oct 2025

Open weights Flagship

No Epoch Capabilities Index score yet

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Benchmark	Score	Rank	Source
MMLU-Pro	82%	#55 of 151	Model card
IFBench	72%	#32 of 70	Model card
Humanity's Last Exam	31.8%	#70 of 147	Model card
LMArena (Elo)	1,344	#181 of 333	Official leaderboard
AA-LCR (Artificial Analysis Long Context Reasoning)	61%	#20 of 31	Model card
Longform Creative Writing (EQ-Bench)	46.4	#90 of 119	Official leaderboard
WideSearch	62.3%	#15 of 17	Model card
Extended NYT Connections	14.8%	#87 of 97	Official leaderboard

Benchmark	Score	Rank	Source
GPQA Diamond	78%	#123 of 284	Model card
AIME 2025	78%	#97 of 143	Model card
SimpleBench	25%	#86 of 100	Official leaderboard

Benchmark	Score	Rank	Source
LiveCodeBench	83%	#9 of 73	Model card
ArtifactsBench	66.8%	#1 of 5	Model card
MLE-bench	40%	#5 of 18	Model card
SWE-bench Verified	61%	#28 of 60	Official leaderboard
Multi-SWE-bench	36.2%	#7 of 10	Model card
SciCode	36%	#39 of 47	Model card
CooperBench	14%	#5 of 6	Official leaderboard
Code Arena	1,297	#106 of 124	Official leaderboard
WebDev Arena	1,298	#91 of 105	Epoch AI
SWE-bench Multilingual	56.5%	#61 of 70	Model card
VIBE-Pro (MiniMax)	42.4%	#3 of 3	Model card

Benchmark	Score	Rank	Source
tau2-bench	87%	#20 of 69	Model card
Elimination Game	5	#19 of 56	Official leaderboard
Terminal-Bench	30%	#8 of 18	Epoch AI
SnakeBench	24	#89 of 183	Official leaderboard
Terminal-Bench Hard	24%	#8 of 11	Model card
Terminal-Bench 2.0	30%	#32 of 43	Official leaderboard
GAIA	75.7%	#4 of 5	Model card
BrowseComp	44%	#65 of 77	Model card
BrowseComp-ZH	48.5%	#15 of 17	Model card
Toolathlon	18.8%	#37 of 37	Model card
xbench-DeepSearch	72%	#3 of 3	Model card

Benchmark	Score	Rank	Source
FinSearchComp	65.5%	#4 of 7	Model card
Physical-Engineering's Last Exam	0.06	#14 of 18	Official leaderboard
Vending-Bench 2	$161	#59 of 66	Official leaderboard
ProfBench	37.7%	#66 of 72	Official leaderboard
TheAgentCompany	36%	#1 of 1	Model card

Benchmark	Score		Rank	Source
Artificial Analysis Intelligence Index v3	61.0		#1 of 1	Model card

The benchmarks the lab chose to headline, with the scores it gave.

Every result in the longer documents the lab published with it. 5 of them aren't in the launch post.