MiniMax M2.7 is live on AI Gateway
By Steven Van ·
The model ships in standard and high-speed variants, the latter running at about 100 tokens per second for twice the cost.
Vercel's AI Gateway now offers MiniMax M2.7, in standard and high-speed variants. Vercel describes it as a major step up from earlier M2-series models for software engineering, agentic workflows, and professional office tasks, with native support for multi-agent collaboration, complex skill orchestration, and dynamic tool search, plus improvements to production debugging and end-to-end project delivery.
The high-speed variant matches standard performance at roughly 100 tokens per second, for twice the cost, intended for latency-sensitive use cases. To use either, set the model to minimax/minimax-m2.7 or minimax/minimax-m2.7-highspeed in the AI SDK.
