Choose speech subtitles or visible hard subtitles
Use speech subtitle generation when the spoken audio should become editable subtitle text.
Use OCR extraction when the subtitle text is already burned into the video frame and no SRT file exists.
Generate and extract subtitles
Start from Subtitle Generator for speech subtitles, or the subtitle extraction API/docs path when automating OCR extraction.
Use a clean source when possible. If old subtitles block readability, consider removing them before creating a final localized video.
Choose the source language if required and decide whether you need SRT, VTT, review text, or translated subtitles.
When the task is ready, download subtitle files from Works or include them in a batch delivery.
Review subtitle files
- Open the SRT or VTT file and scan for missing lines, repeated text, or timing gaps.
- Check names, brand terms, product names, and numbers before translation.
- Watch a short section with subtitles enabled to confirm timing and line breaks.
- Use the subtitle file as source text for translation, dubbing, QA, or a downstream editor.
Visible hard subtitles that are blurry, tiny, partially covered, or low contrast can reduce OCR accuracy.
Automate repeated subtitle work
Use API automation when subtitle generation or extraction is part of a larger CMS, archive conversion, media operations, or customer delivery workflow.