MiniMax · Language model

MiniMax M3.1 Flash

MiniMax's cheap reasoning model for coding and agent work. MiniMax says the weights follow on October 23.

MakerMiniMax
ReleasedOctober 19, 2026
API price (per 1M tokens)$0.15 in / $0.6 out
Open weightsNo

Best for

  • Low-cost coding agents
  • High-volume jobs where you check every answer
  • Experiments ahead of the open-weight release

Strengths

  • Scores 42–43 on the Artificial Analysis Intelligence Index, up 13 points on MiniMax M3
  • About $0.22 per Artificial Analysis index task, below GLM-5.3-Flash ($0.28) and DeepSeek V4.1 Flash ($0.26)
  • Cache reads at $0.015 per million tokens

Watch out for

  • 93% hallucination rate on Artificial Analysis's AA-Omniscience test: it almost never says 'I don't know'
  • Open weights are promised but not yet out, and the license isn't published

Ranking history

EditionPower RankingsCodingBest ValueOpen-Weight WatchImage GenerationVideo Generation
Week 45, 2026——#8 (85)———
Week 44, 2026——#7 (87)———