XiaomiMiMo V2.5 Speech
MiMo V2.5 TTS VoiceClone
VoiceClone generates speech using a voice sampled from audio. Style can still be controlled with text instructions and audio tags.
Choose reference audio for the result you need
Practical guidance. Use an authorized recording with clear speech and little background noise. Check whether the output preserves pronunciation and voice identity across different sentences. A short successful sample is not enough to judge long narration or a different language.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | mimo-v2.5-tts-voiceclone |
| Sample input | The documented route takes MP3 or WAV audio as Base64, with a 10 MB encoded-size limit.MiMo V2.5 speech synthesis guide |
| Separate capability | The current catalogue identifies voice cloning as a separate speech model, not a preset-voice option.MiMo current model catalogue |
Related models
- MiMo V2.5 TTS
Xiaomi · MiMo V2.5 Speech
The standard V2.5 TTS model turns text into speech using Xiaomi's built-in voices.
Launch date not confirmedAPI - MiMo V2.5 TTS VoiceDesign
Xiaomi · MiMo V2.5 Speech
VoiceDesign creates a new voice from a written description.
Launch date not confirmedAPI