MiniMax releases M3.1 Flash, a low-cost model for agents and coding
MiniMax rolled out M3.1 Flash, a faster, cheaper model aimed at agent tasks, coding, computer use and security work, priced at $0.15 per million input tokens and $0.60 per million output tokens through its API.
What happened
- On October 19, 2026, MiniMax officially released M3.1 Flash and said an early version had already run anonymously as "Space Bunny".
- MiniMax says online inference is nearly three times faster and overall agent-task cost is 70% lower than the previous Flash model.
- Pricing is $0.15 per million input tokens, $0.60 per million output tokens and $0.015 per million cached input tokens.
- MiniMax reports scores including 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2 and 82% on CyberGym.
- It's available now through the MiniMax API, the M Plan subscription and MiniMax Code; MiniMax says model weights and a Pro model will come later.
Why it matters: Cheaper models that can handle multi-step agent work lower the cost of automating routine business tasks. Benchmark figures are MiniMax's own and haven't been independently confirmed.
Sources
- MiniMax: MiniMax M3.1 Flash minimax.io
- SiliconANGLE siliconangle.com
Checked October 20, 2026 against the sources above. Claims about performance or benchmarks are the company's own unless stated otherwise.