MiniMax launches M3.1 Flash at $0.15 per million input tokens
Original titleMiniMax Just Launched M3.1 Flash: 428B MoE Agentic Model at $0.15/M Input
AISummary
MiniMax has released M3.1 Flash on its API at $0.15 per million input tokens and $0.60 per million output tokens. The 428B-parameter mixture-of-experts model has about 23B active parameters and a 1,000,000-token context window. MiniMax reports leading Automation-Bench v1.0.6 at 62.2 but trailing Opus 5 on Terminal Bench 4.0, and model weights are due October 23.
This story is featured in top picks: “MiniMax releases M3.1 Flash with faster inference and 70% lower agent costs”
Source: MarkTechPost · marktechpost.comPublished · added here