An MP3 voiceover and a timed SRT subtitle file in your project.
About
Turns a script into a voiceover. It writes an MP3 into your project and, next to it, an SRT subtitle file timed to the speech. You can choose the voice (including one made with a voice design tool), the speech model, and how steady and how close to the original voice it sounds; the brand's saved voice settings are used when you set none.
By default the audio is loudness-normalised to a standard spoken-word level (-16 LUFS) so it sits correctly against music. This step needs FFmpeg, and if it fails the raw audio is kept.
Each voiceover is a paid call to a speech model, billed by the number of characters. In Syvon a typical 1,000 characters cost about $0.42, or about $0.21 on the Flash model.
- TypeScript
- AI speech models
- FFmpeg
Requirements
- Runs with
- Node.js 20 or later
- Operating systems
- windows, macos, linux
- Needs installed
- FFmpeg
- Network
- Sends the text to the speech provider over the internet.
Costs
- External costs
- ElevenLabs: billed to your own key — Planned direct lane: you would bring your own ElevenLabs key. Today the call runs through the Syvon provider proxy and Syvon credits.