Caption reliability layer
Fix captions for delivery — see what broke and what we fixed.
Drop in an .srt, .vtt, .sbv, .ass/.ssa, .ttml/.dfxp/.xml, or CaptionPass .json IR file. Pick the platform you’re shipping to. CaptionPass applies a deterministic rule pipeline, returns a clean file, and shows you every fix it made and every issue it could not silently repair.
Drop an .srt, .vtt, .sbv, .ass/.ssa, .ttml/.dfxp/.xml, or .json file here, or click to browse.
We’ll show a validation report here once you upload. Every fix is named and traceable to its cue.
- Format and timestamp normalization
- Overlap and reading-speed checks
- Platform-specific styling stripping
Explainable fixes
Every transformation is named and traceable to a cue, so you can audit before re-uploading.
Platform-aware presets
Line length, reading speed, and styling rules tuned for YouTube, TikTok, HTML5, LMS/TTML, a generic safe baseline, and a developer JSON IR for automation.
Ephemeral by default
Files are decoded in memory for the request and never persisted. No accounts needed for the free tier.