From the source
MiniMax has released M3.1 Flash, a reasoning model that scores 42 points on the Artificial Analysis Intelligence Index, 13 points above its predecessor M3 at less than half the cost per task.
- The model ties with GLM-5.3-Flash, edges past Gemini 3.8 Flash (41) and trails
Deepseek V4.1 Flash 1017 (45).
- According to Artificial Analysis, it hits 48 percent on Humanity's Last Exam, 22 percent on the CritPt research physics benchmark and 56 percent on the AA-Omniscience factual-accuracy test, but shows a 93 percent hallucination rate because it answers every question rather than declining when uncertain.
- The model handles up to one million tokens of context and accepts text and images.
- API pricing is $0.15 per million input tokens and $0.60 per million output tokens, with cached inputs at $0.015, and the API is live at platform.minimax.io.
- MiniMax says M3.1 Flash runs three times faster than M3 and plans to release the model as open weights on October 23.