News
No articles yet.
Models
Minimax M1 40k
MiniMax
Context40K
Input/1M$0.4
Minimax M1 80k
MiniMax
Context80K
Input/1M$0.4
MiniMax M3
MiniMax
Context1000K
Jun 12, 2026
MiniMax M2.7
MiniMax
Context205K
Input/1M$0.3
Mar 16, 2026
MiniMax-M2.5
MiniMax
MiniMax's latest open-weight reasoning model with improved coding and agentic task performance.
Context197K
Input/1M$0.3
Feb 12, 2026
Open weightsMiniMax-M2
MiniMax
230B reasoning model with 1M context, MIT license. Top-ranked open-weight model on Artificial Analysis.
Context1000K
Input/1M$0.3
Oct 26, 2025
Open weightsMiniMax-M1
MiniMax
Open-weight hybrid MoE reasoning model with 456B parameters and 1M-token context.
Context1000K
Input/1M$0.4
Jun 17, 2025
Open weightsMiniMax-Text-01
MiniMax
456B MoE model with Lightning Attention enabling a native 1M-token context window. Open-sourced.
Context1000K
Input/1M$0.2
Jan 15, 2025
Open weightsTop Benchmark Scores
Full leaderboard →AIME 2025
MiniMax-M2Arena Elo
MiniMax M3GPQA
MiniMax M3Hallucination Rate
MiniMax-M2.5LiveCodeBench
MiniMax-M2MMLU-Pro
MiniMax-M2MMMU
MiniMax M3Speed (tok/s)
MiniMax M3SWE-bench Verified
MiniMax M3