TL;DR: MiniMax released M3.1 Flash, a new model offering frontier-level intelligence at significantly reduced cost and increased speed, redefining performance economics for agentic AI.
Summary: MiniMax has launched its M3.1 Flash model, an advanced version of their Space Bunny model. M3.1 Flash achieves high performance across tasks like Cowork (Automation-Bench: 62%; ALE: 30%), Coding (DeepSWE: 74%), and Visual Reasoning (BabyVision: 91%). It operates at one-third the cost and three times the speed of its predecessor, M3.
Why it matters: This release provides AI developers with a more cost-effective and faster model for building agentic AI applications. Builders should explore M3.1 Flash for projects requiring high intelligence at improved economic efficiency, especially for automation, coding, and visual reasoning tasks.
Source: x_com