MiniMax M2
minimax/minimax-m2Efficient open reasoning tuned for agent loops, low price, interleaved thinking, dependable tools.
200K ctx · Free
稀宇科技MiniMax
Multimodal breadth: text, video, speech
Models
4
Founded
2021
Country
China
Shanghai
MiniMax has been building models out of Shanghai, China, since 2021. Webparam tracks four of its models, spanning reasoning, language, video and speech: free to call and reaching 1M. Two of them ship with .
Models
The MiniMax catalogue
Ordered by weekly usage on Webparam.
MiniMax M2 by MiniMax MiniMaxOn request MiniMax Text 01 by MiniMax MiniMaxOn request MiniMax Text 01
minimax/minimax-text-01Million-token open language model, lightning-attention architecture built for very long inputs.
1M ctx · Free
Hailuo Video 02 by MiniMax MiniMaxOn request Hailuo Video 02
minimax/hailuo-video-02Cinematic short-form video generation with notably natural human motion.
— ctx · Free
Speech 02 by MiniMax MiniMaxOn request Speech 02
minimax/speech-02Production text-to-speech with expressive multilingual voices, priced per character.
— ctx · Free
Recent releases
What shipped lately
History
MiniMax builds across more modalities than almost any peer, million-token text models, the Hailuo video line, and production-grade speech synthesis. Its M-series brought efficient open reasoning to the mix, rounding out one of the most complete catalogues in the ecosystem.
MiniMax builds across more modalities than almost any peer, million-token text models, the Hailuo video line, and production-grade speech synthesis. Its M-series brought efficient open reasoning to the mix, rounding out one of the most complete catalogues in the ecosystem.
Strengths
- True multimodal catalogue: text, video and speech
- Million-token context on the text line
- Competitive open reasoning models
Use MiniMax models through Webparam
Every model above is callable through one OpenAI-compatible convention: change the model ID, keep the request.
Related providers