MiniMax's M3.1 Flash cuts agent task costs by 70% while running nearly 3x faster
MiniMax M3.1 Flash
MiniMax has officially launched M3.1 Flash, the model behind the anonymous "Space Bunny" that processed over 200 trillion tokens. It matches leading models on agentic tasks, coding, computer use, and cybersecurity benchmarks, including 73.7% on DeepSWE v1.1 and 82% on CyberGym. Inference runs nearly three times faster, with output speeds above 200 tokens per second, and pricing is set at $0.15 per million input tokens and $0.60 per million output tokens.
Space Bunny, which processed more than 200 trillion tokens while operating anonymously, was an early version of MiniMax M3.1 Flash.