MiniMax launches M3.1 Flash: frontier performance at 1/3 the cost of M3
aziz4ai · x · 2026-10-21
MiniMax officially launched M3.1 Flash — the formal version of "Space Bunny", which topped global leaderboards with 200T+ tokens processed — promising frontier-level intelligence at radically lower cost.
Key results: Cowork (Automation-Bench 62%, ALE 30%), Coding (DeepSWE 74%), and Visual Reasoning (BabyVision 91%), at 1/3 the cost of M3 and 3x the speed. Flat API pricing: $0.15 / $0.60 per 1M input/output tokens, cache reads at $0.015. Model weights arrive Oct 23; available via Runware, the API, M Plan, and the MiniMax Code desktop app.
Related event: MiniMax launches M3.1 Flash at one-third the price of M3(3 posts)→
More from Models
- Anthropic's free usage limit reset expires tomorrow at 9am PT — lydiahallie · 2026-10-22
- ARC-AGI-3: Tong Hui Kang hits 20th, a whisker from gold; Nader enters ARC-2 gold — Youssef_M_Nader · 2026-10-22
- Anthropic launches Duet, an always-on personal agent that shifted this tester from ChatGPT to 50/50 Claude — petergyang · 2026-10-22
- DeepSeek has open-sourced every layer of an AI lab, from weights to inference engine — examachine · 2026-10-22
- Anthropic's proactive agent Claude Duet spotted: internal name "aibo", Stripe shopping demo — testingcatalog · 2026-10-22
- Recursive launches Flint, a laptop coding model largely built by AI — TechCrunch AI · 2026-10-22