Benchmark Almanac

MiniMax-M3.1-Flash

MiniMax-M3.1-Flash places in the top three on 2 benchmarks.

MiniMax Released 19 Oct 2026 Family MiniMax M3 The lab's model card
33
Benchmarks with a score
0
Ranked first, out of five or more models
2
In the top three
MiniMax Oct 2026

MiniMax-M3.1-Flash

Mid-size

No Epoch Capabilities Index score yet
  • NL2Repo-Bench 61.2% #4 /28
  • Agents' Last Exam 29.5% #10 /12
  • Legal Research Bench 34.1% #47 /88

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.

Knowledge and language

BenchmarkScoreRankSource
Humanity's Last Exam 43.4% #53 of 147 Model card
LOCA-Bench 60.3% #1 of 2 Model card
Extended NYT Connections 28.3% #77 of 97 Official leaderboard

Coding

BenchmarkScoreRankSource
NL2Repo-Bench 61.2% #4 of 28 Model card
DeepSWE 73.7% #12 of 51 Model card
Code Arena 1,582 #33 of 124 Official leaderboard
Vibe Code Bench 73.9% #41 of 115 Official leaderboard
PostTrainBench 1.1 39.3% #1 of 2 Model card
OpenDesign Arena 81.7 #10 of 19 Official leaderboard
SWE Refactor Bench 23% #1 of 1 Model card

Agents and tools

BenchmarkScoreRankSource
AutomationBench 62.8% #2 of 41 Model card
OSWorld 2.0 61.7% #9 of 19 Model card
Terminal-Bench 4.0 38.1% #15 of 23 Model card
Agent Arena -0.04 #36 of 53 Official leaderboard

Professional work

BenchmarkScoreRankSource
FoodTruck Bench $63,464 #2 of 10 Official leaderboard
Tax Agent Bench 32.1% #20 of 79 Official leaderboard
Legal Research Bench 34.1% #47 of 88 Official leaderboard
Vals Finance Index 55.9% #21 of 35 Official leaderboard
Excel Modeling Benchmark v1.1 67.8% #24 of 37 Official leaderboard
Vals Index v2.2 53.8% #21 of 32 Official leaderboard
BankerToolBench 76.2% #2 of 3 Model card
Harvey LAB v1.1 (HLAB-Vals-1.1) 7.50% #23 of 33 Official leaderboard
Vals Legal Index 67.1% #24 of 33 Official leaderboard
Vals Finance Agent v2.1 58.8% #26 of 35 Official leaderboard
Agents' Last Exam 29.5% #10 of 12 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
BabyVision 91% #4 of 20 Model card
CharXiv 82.5% #20 of 67 Model card
VoxelBench 1,799 #23 of 56 Official leaderboard
GDP.pdf 20.8% #13 of 15 Model card

Safety and security

BenchmarkScoreRankSource
CyberGym 82.2% #13 of 33 Model card
FORTRESS 52.1% #53 of 66 Official leaderboard

Other

BenchmarkScoreRankSource
LiveBench 66.9% #16 of 28 Official leaderboard
SAGE v1.1 (Vals) 85.3% #17 of 28 Official leaderboard

Its launch post

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