MiniMax M3.1 Flash
MiniMax has unveiled its latest model, MiniMax M3.1 Flash, a next-generation Flash model with a significant leap in agentic tasks, coding, computer use, and cybersecurity. This new model is three times faster in online inference and costs 70% less for usage. M3.1 Flash achieves performance levels similar to leading models in complex agentic tasks, computer use, software development, and cybersecurity.
The model's inference efficiency has been optimized through sparsity, achieving output speeds of 200+ TPS (tokens per second) and median speeds of 170+ TPS. Pricing is flat per 1M tokens: $0.15 input, $0.60 output, and $0.015 cached input. M3.1 Flash excels in Automation-Bench, DeepSWE, and Agents' Last Exam benchmarks, offering performance comparable to top models at a lower cost.
The model is accessible via MiniMax API, M Plan, and MiniMax Code. Model weights will be released later.
Written by urgent.news from Hacker News Best's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- MiniMax M3.1 Flash minimax.io
