MiMo TTS

Voice design meets cloning.

Xiaomi MiMo-V2.5-TTS with preset voices, text-based voice design, and audio-based voice cloning. Built for Chinese and English.

Model
MiMo-V2.5
Voices
9 presets (5 CN, 4 EN)
Modes
3 modes
Languages
CN + EN optimized
Capabilities

Three ways to create voices

01

Preset voices

9 handcrafted voices: 冰糖, 茉莉, 苏打, 白桦, Mia, Chloe, Milo, Dean, and a default.

02

Voice design

Describe a voice in text. MiMo synthesizes it — no audio sample needed.

03

Voice clone

Upload audio and MiMo creates a clone that preserves the original voice characteristics.

FAQ

Common questions

How does MiMo compare to ElevenLabs?

MiMo is optimized for Chinese with native voice design. ElevenLabs leads in English multilingual. Voicebench lets you use both.

What languages does MiMo support?

Chinese and English with high fidelity. Other languages are in development.

Get started

Ready to try every voice?

Generate, clone, narrate and broadcast from a single workspace. No credit card required to start.