Private AI
MiniMax-M1 is a hybrid MoE reasoning model with 40K thinking budget. World's first open-weight, large-scale hybrid-attention model with lightning attention for efficient test-time compute scaling. Excels at complex tasks requiring extensive reasoning.
Added Jan 8, 2025
Context Window
1.0M
Max Output
131.1K
Input Price (Auto)
$0.14/1M
Output Price (Auto)
$1.33/1M
Performance metrics and benchmarks
Sourced from LMArena.
Arena Score
1363.6
Overall Rank
#151 / 360
Votes
35,214
Confidence Interval
1359.4 - 1367.9
Coding
#146 / 355
6,489 votes
1416.0
Math
#146 / 349
1,799 votes
1370.8
Longer Query
#156 / 338
6,963 votes
1364.6
Creative Writing
#164 / 358
4,546 votes
1317.7
Instruction Following
#158 / 360
8,684 votes
1345.6
Hard Prompts
#151 / 360
15,829 votes
1380.7
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
French
#108 / 245
425 votes
1411.7
German
#127 / 270
828 votes
1359.4
Industry Mathematical
#138 / 342
1,785 votes
1381.6
Industry Medicine And Healthcare
#142 / 329
2,116 votes
1393.2
English
#145 / 360
16,873 votes
1384.8
Spanish
#147 / 248
716 votes
1353.5
Industry Life And Physical And Social Science
#148 / 358
5,885 votes
1383.4
Chinese
#149 / 330
1,822 votes
1385.8
Industry Software And It Services
#150 / 360
11,668 votes
1402.4
Exclude Ties
#152 / 360
24,825 votes
1329.3
Hard Prompts English
#154 / 359
8,158 votes
1394.3
Multi Turn
#154 / 358
5,695 votes
1357.3
Expert
#155 / 309
1,594 votes
1363.8
Industry Entertainment And Sports And Media
#155 / 358
6,296 votes
1325.2
Russian
#158 / 323
1,930 votes
1348.3
Industry Business And Management And Financial Operations
#159 / 353
6,116 votes
1352.0
Korean
#159 / 240
635 votes
1272.4
Industry Legal And Government
#165 / 333
2,408 votes
1365.1
Japanese
#165 / 228
773 votes
1225.3
Industry Writing And Literature And Language
#166 / 359
7,800 votes
1331.6
Published 2026-05-27 · Matched as minimax-m1
LMArena Dataset