MiniMax M3.1 Flash: the mystery 'Space Bunny' model officially launches
MiniMax officially released M3.1 Flash, revealing it as the anonymous 'Space Bunny' model that developers had been testing for two weeks; $0.60 per million output tokens with Automation-Bench scores above Opus 5.5, weights open-sourcing October 23.
Why it matters
Two weeks of anonymous real-world testing before the launch gave it organic credibility most releases lack, and its cost-per-task is roughly 1/20th of frontier rivals on business-automation benchmarks.
What you could build with it
A startup building back-office or customer-support automation agents could route multi-step tool-calling workflows through M3.1 Flash instead of flagship models, keeping agent-task completion near frontier levels at roughly 1/20th the cost, for example an agent that processes invoices, updates ERP records and drafts follow-up emails.
Does it hold up?
Too early to judge: launched today, and while the two-week anonymous 'Space Bunny' run gave it real-world exposure, no widely published independent benchmark scores exist yet.
Built with MiniMax M3.1 Flash
- StartupFortune: M3.1-Flash launched quietly after weeks as mystery 'Space Bunny'article · Analysis of MiniMax's stealth-launch strategy: Space Bunny let developers stress-test the model for two weeks before the maker claimed it, plus early private-eval caveats.
- MiniMax M3.1 Flash cuts agent task costs by 70%, Pro version in pipelinearticle · Writeup on the release: API/M Plan/MiniMax Code availability, weights open-source soon, and limited-time MiniMax Code offers tied to the launch.
- Japanese walkthrough: M3.1 Flash benchmark figures, pricing and usabilityyoutube · Video explainer comparing M3.1 Flash's official benchmark figures against Opus 5.5 and DeepSeek-V4.1-Flash-1017 on cost-per-score graphs.
- MiniMax M3.1 Flash weights go public on HF/ModelScope; day-0 on Together AIkie.ai · Oct 23 update: 428B total / 23B active weights published on Hugging Face and ModelScope under the MiniMax Community License; Together AI, Runware and Inferact hosted day-0; community quants for llama.cpp, Ollama and LM Studio.
Learn more
First spotted on article: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73NVIDIA Isaac Lab 3.0 GA + Isaac OS open runtime + Agile ONE SN humanoidMajor Isaac platform expansion at GTC Berlin Oct 21: Isaac Lab 3.0 is generally available as the open-source robot…models · JEV 0.73StepFun Step-5-Preview — open BF16 weights releasedStepFun published the full BF16 weights for its flagship Step-5-Preview agentic model (600B-parameter sparse MoE, 27B…models · JEV 0.73NVIDIA Nemotron 3.5 Retriever collection: Embed 8B/1B, Rerank 8B, Parse 631MNVIDIA's Oct 21 collection drop for agentic retrieval: Nemotron-3.5-Embed in 8B BF16, 1B BF16 and 1B NVFP4…models · JEV 0.72Kimi K3.1 Pro: Moonshot's 2.8T open-weight flagship officially launchesMoonshot AI officially launched Kimi K3.1 Pro on Oct 23, a mixture-of-experts model with 2.8 trillion total / 104…models · JEV 0.72
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.