中文
Back to the feed

Xiaomi MiMo text-to-speech API is free for now, voice cloning included

#MiMo#Xiaomi#TTS#API deals

Xiaomi made its three MiMo TTS models free for a limited time: standard speech, voice cloning, and voice design. Calls don't charge your pay-as-you-go balance or use Token Plan credits. Sign up, create an API key, and call them through an OpenAI-compatible endpoint.

The briefing

Xiaomi has made the MiMo TTS series free for a limited time. Calls with a pay-as-you-go API key don't charge your balance, and calls on a Token Plan don't use plan credits. If you want voiceovers, audio content, or speech in your own app, you can start for free now.

What you can use

  • mimo-v2.5-tts is the standard speech model. It comes with built-in voices, and you can steer tone, emotion, and pace with a one-line or longer style prompt.
  • mimo-v2.5-tts-voiceclone clones a voice. Send an audio sample and it reads new text in that voice.
  • mimo-v2.5-tts-voicedesign builds a voice from a text description.
  • Each request handles up to 8K tokens at 100 requests per minute. Output can be wav, mp3, or pcm, with streaming support.
  • It works on both the China and international platforms. Speech recognition and the language models still bill as usual; only the TTS series is free.

How to use it

  1. Sign up on the Xiaomi MiMo platform and create an API key in the console.
  2. Send requests to https://api.xiaomimimo.com/v1/chat/completions with the model set to mimo-v2.5-tts, mimo-v2.5-tts-voiceclone, or mimo-v2.5-tts-voicedesign. You can also point the OpenAI SDK's base_url at https://api.xiaomimimo.com/v1.
  3. Put the text to speak in an assistant message, the style prompt in a user message, and pick the format and voice in the audio parameter.

See the MiMo-V2.5-TTS model page, the TTS API docs, and pricing.

Is this still live?

Community signal100% say it still works