Video and voice
Voice-over by speech synthesis
Enter some text, choose a timbre: Studio produces a synthetic voice you can download as WAV or lay over a video.
The voice is generated by an AI model (SpeechT5) trained on English: French is read with an English accent.
- Free
- No sign-up
- Text and audio stay on your device
This tool uses an artificial intelligence model that runs in your browser. The resulting voice is a synthetic voice. Learn more about how AI is used
The voice model (SpeechT5) is trained on English: it also reads French, but with an English-speaking accent. The voices below are real timbres — the result is a real audio file, free and generated on your device.
Crisp female timbre. English-language model: English accent on French text.
The voice model is downloaded from a third-party server (Hugging Face) the first time you run it, then runs in your browser. That server receives your IP address, never your text. A synthetic voice must be presented as such when you distribute it. How Studio uses AI.
Missing the video?
Create a short clip from a photo, then come back here to add the narration.