- New model: MiniMax-M3.1-Flash. A multimodal reasoning model for coding and agent work: reads text, images and video, 1M-token context, function calling, thinking always on. Call it as
MiniMax-M3.1-Flashfrom the API, your coding tools, Koda, or pick it in Chat. - Free for every plan until the end of Oct 23 (IST). Requests to it use none of your quota while the offer runs, free plan included. A fair-use limit of 10 requests a minute per account keeps it fast for everyone. After the offer it becomes a regular Pro-tier model at 1× a request.