MiniMax launches M3.1 Flash model with 3x faster speed and 70% lower cost

MiniMax has officially launched its new M3.1 Flash AI model. Space Bunny, an anonymous model that handled over 200T tokens, was actually an early version of it. The new model brings major upgrades in agent tasks, coding, computer use, and cybersecurity, with nearly 3x faster inference and 70% lower costs.

In benchmark tests, it scored 73.7% on the DeepSWE coding test, 82% on the CyberGym security test, and 61.7 on OSWorld V2 for computer use, ranking first among AI models developed in China.

It costs 1 yuan per million input tokens and 4 yuan per million output tokens. The model is now live on MiniMax API, MPlan, and MiniMax Code, with open weights coming soon, and a more powerful Pro version on the way.