MiniMax
Use MiniMax when an agent needs MiniMax or Hailuo multimodal AI: M3 and M2.7 text, chat, responses, and token estimates; speech-2.8 or speech-2.6 short and asynchronous long-form text-to-speech; voice cloning, design, and management; Hailuo 2.3 text-, image-, or reference-to-video and Video Agent template jobs with status polling; image-01 generation; music-3.0 songs, lyrics, and cover preprocessing; or model and file listing, upload, retrieval, and deletion. Choose it for MiniMax-specific models or one provider spanning text, audio, image, video, and music; use another integration when the user requests a different provider. A MiniMax API key is required. Ask before paid generation, uploading private media, cloning or designing voices, or deleting files or voices. Platform credentials are limited to reviewed bounded-cost stateless actions and safe public metadata or status reads.