Paste a raw transcript from YouTube, Zoom, Teams, Otter or any captioning tool. Timestamps, speaker labels, filler words such as um and uh, stutters and [inaudible] notes are removed, and choppy lines are joined into readable paragraphs.
Drop a text file
0words before
0words after
0fillers removed
0timestamps removed
Advertisement
Advertisement
How to use the Transcript Cleaner
Paste your transcript or open a .txt, .srt or .vtt file. An example loads automatically.
Choose what to remove: timestamps, speaker labels, filler words, stutters and bracketed notes.
Pick a layout: paragraphs (one per speaker turn), the original line breaks, or one sentence per line.
Add your own filler words, separated by commas, if your speaker has a favorite one.
Copy or download the clean transcript.
What gets cleaned
Problem
Raw transcript
After cleaning
Timestamps
0:05 so today we're going to 0:07 talk about pricing
So today we're going to talk about pricing
Bracketed times
[00:01:23] Thanks for coming.
Thanks for coming.
Speaker headers
Speaker 2 0:15 We shipped it.
Speaker 2: We shipped it.
Filler words
Um, so, uh, I think it's, um, ready.
So, I think it's ready.
Stutters
I I think the the m- main issue
I think the main issue
Non-speech notes
It was [inaudible] fine. (laughs)
It was fine.
Broken lines
we saw a big jump in sign-ups
We saw a big jump in sign-ups
Verbatim vs. clean verbatim
Professional transcription uses two main styles. Strict verbatim records every sound: fillers, false starts, repetitions, laughter and pauses. It is used in legal proceedings and research where how something was said matters. Clean verbatim keeps the speaker's words and meaning but removes the noise, which makes it the usual choice for blog posts, meeting notes, show notes, captions and quotes.
This cleaner turns raw or strict verbatim text into clean verbatim. It never rewrites sentences or changes word choice, so the result is still faithful to what was said. Always proofread quotes before publishing: automatic captions often mishear names and technical terms.
Tips for common sources
YouTube: open the video description, click Show transcript, select all lines and paste. Timestamps and the "2 minutes, 5 seconds" lines disappear. For the caption file itself, use SRT to Text.
Zoom and Teams: paste the .vtt or .txt transcript. Speaker names are kept and consecutive turns by the same person can be merged.
Otter and similar apps: the "Name 0:03" header lines become "Name:" labels at the start of each paragraph.
Frequently asked questions
How do I remove timestamps from a transcript?
Paste the transcript and keep Remove timestamps ticked. The cleaner removes stand-alone time lines such as 0:05 or 12:41, bracketed times like [00:01:23] or (2:15), times at the start of a line, SRT and VTT timing lines with -->, and the spoken-length lines YouTube adds such as 1 minute, 5 seconds.
Which filler words are removed?
By default: um, umm, uh, uhh, uhm, er, erm, ah, hmm, mm, mhm, mm-hmm and uh-huh. Tick the verbal tics option to also remove like, you know, I mean, sort of, kind of, basically, actually and literally, but only where they are set off by a comma, so normal sentences such as I like it stay untouched. You can add your own words too.
What is a clean verbatim transcript?
A clean verbatim (or intelligent verbatim) transcript keeps everything the speakers said but leaves out fillers, false starts, stutters and non-speech sounds, and adds normal punctuation. That is what this tool produces. A strict verbatim transcript keeps every um and repetition.
How are speaker labels detected?
Labels such as John:, Speaker 1:, SPEAKER_02:, Interviewer: or >> at the start of a line are recognized, as well as Otter and Teams style header lines with a name followed by a time. A capitalized name is only treated as a label when it appears at least twice, to avoid removing ordinary text like Note:.
Can it fix the line breaks in a YouTube transcript?
Yes. YouTube's Show transcript panel copies every few words on a new line with a timestamp. With Paragraphs selected, the lines are joined into sentences and grouped into paragraphs of about five sentences, or one paragraph per speaker.
Is my transcript private?
Yes. All cleaning happens in your browser. Your text is never uploaded or stored on a server.