- Text tab → Captions → choose the spoken language and quality (Fast ≈ 40 MB, Accurate ≈ 80 MB), then Generate captions.
- Speech-to-text runs on your computer with OpenAI's Whisper model. Your audio is never uploaded. The model downloads once and is reused.
- Edit any line, delete lines, choose Dark box / Outline / Yellow, and top or bottom position. Lines also appear on the timeline's Captions row, where you can retime them.
- Download .srt for YouTube and other players. Its timings follow your cuts and speed changes.
- Home
- Help center
- Captions
Help · Editor
Captions
Text tab → Captions → choose the spoken language and quality ( Fast ≈ 40 MB, Accurate ≈ 80 MB), then Generate captions .