MiniMax launches M3.1 Flash: 1/3 the cost, 3x speed of M3, weights open on Oct 23
woosuk_k · x · 2026-10-20
MiniMax releases M3.1 Flash
MiniMax officially launched M3.1 Flash, confirming that "Space Bunny" — the model that topped global leaderboards while processing 200T+ tokens — was an earlier version of it. Positioning: ultra-efficient agentic model.
- Key scores: AutomationBench 62%, ALE (Agents' Last Exam) 30%, DeepSWE 74%, BabyVision visual reasoning 91%
- Economics: frontier-level intelligence at 1/3 the cost of M3 and 3x the speed
- API pricing (flat rate): $0.15 input / $0.60 output per 1M tokens; cache reads at $0.015/1M
- Specs: 1M-token context window, tunable thinking depth via effort, aimed at agent reasoning, tool use, coding and long context
Weights and ecosystem
Open weights arrive Oct 23. Inference provider Inferact is the API launch partner, contributing its vLLM engineering and platform for production serving. MiniMax Code desktop app and M Plan subscription are also available.
More from Models
- REAX inference engine runs Perplexity Decider up to 6x faster, free for individuals — airesearch12 · 2026-10-20
- SKY-001 debuts as first 'Engineering World Model,' claiming 125x speedup over GPT-6 on CAD — vasuman · 2026-10-20
- Smart glasses test: Octen search at 60ms, model generation is the bottleneck; Mercury 3 wins — Scobleizer · 2026-10-20
- BitNet CPU inference is silently wrong: perplexity 6x reference, one-line fix works — matheus-c-feitosa · 2026-10-20
- Space Bunny stealth model revealed as MiniMax M3.1 Flash: 428B-A23B, weights Thursday — BTA_Labs · 2026-10-20
- Clairic is live on Workers AI; open-weights release planned once the architecture matures — michellechen · 2026-10-20