MiniMax M2.1 now live on Vercel AI Gateway
By Steven Van ·
MiniMax's M2.1 model is now callable through Vercel AI Gateway as minimax/minimax-m2.1, with no separate provider account needed.
MiniMax's M2.1 model is now available through Vercel's AI Gateway, with no separate MiniMax provider account required. The model is faster than its predecessor, M2, with improvements in coding tasks and complicated multi-step tool-calling workflows: higher quality code, better instruction-following on difficult tasks, and a cleaner reasoning process, plus gains across multiple languages including Go, C++, JS, C#, and TS, and across refactoring, feature adds, bug fixes, and code review.
To use it via the AI SDK, set the model to minimax/minimax-m2.1. As with other AI Gateway models, requests get unified usage and cost tracking, retries, and provider failover.