Voice Studio

Paste text, choose a voice and tone, and get natural speech for shadowing practice.

Gemini + OpenAI voicesTone and emotion controlOne TTS request per chunk

Model provider

Voice

API key Your text Tone and emotion (describe how it should sound)

Simple mode: each generation uses one Gemini TTS request. This helps you conserve your free API quota.

Advanced: model names

Speech model Text model (for content-aware step, Gemini only) If you get a "model not found" error, check the provider's docs for the current name and paste it here.