What does the transcript formatter do?

It takes raw machine-generated captions (Whisper output, YouTube auto-captions, SRT/VTT files) and produces clean readable text: no timestamps, no cue numbers, broken lines merged into sentences, and paragraphs restored at punctuation boundaries.

Does it work with YouTube auto-generated captions?

Yes. Copy the caption text (with or without timestamps) from YouTube's transcript panel and paste it in. The formatter detects and strips the timing metadata automatically.

Transcript Formatter — Clean Up Raw Whisper & YouTube Captions

Paste raw Whisper output, YouTube auto-captions or any SRT/VTT — get a clean, readable transcript with real sentences and paragraphs.

Why machine transcripts look broken

Whisper, YouTube auto-captions and other ASR systems produce text optimized for timing, not reading: each cue is a short fragment, lines break mid-sentence, and punctuation is inconsistent or missing. Reading a raw transcript is exhausting.

What this formatter fixes

Strips timestamps, cue numbers and WEBVTT headers
Merges fragmented caption lines back into complete sentences
Rebuilds paragraphs at sentence boundaries (configurable length)
Cleans ASR artifacts: repeated [Music]/[Applause] markers (optionally kept), speaker-line heuristics
Collapses duplicated words left over from rolling captions

Perfect for

Turning a 1-hour podcast transcript into readable show notes
Cleaning up lecture or meeting recordings transcribed with Whisper
Prepping YouTube auto-captions for blog publication

Transcript Formatter

Why machine transcripts look broken

What this formatter fixes

Perfect for

FAQ