Overtone
Overtone cuts captions separately for wide and vertical video.
- AI tool
- Free
Pricing
Free and open source (MIT); transcription uses your own provider keys.
Overview
Overtone is an open-source (MIT) caption toolkit that records every word a single time along with its timing and speaker, and then lays out caption lines separately for each screen shape, for example widescreen and vertical phone video. Line breaks follow reading rules, for example avoiding a line that ends on 'the' and keeping names together. A command-line tool can re-cut existing SRT or WebVTT files offline, or create captions from audio using transcription and timing services, with a language model fixing names but never timestamps. A plugin for the video.js player picks whichever cut best fits the picture.
Key features
- Separate line breaks for each screen shape
- Re-cut existing SRT/WebVTT offline
- Transcribe from audio (OpenAI, Azure, ElevenLabs)
- video.js plugin
- Exports to WebVTT, SRT and text
Use cases
Alternatives
Last updated 2026-11-01