How this calculator works
Timestamps in common formats are removed, including 0:00, 00:00:00, [00:12] and subtitle ranges such as 00:00:01,000 --> 00:00:04,000. Line numbers from SRT files and the WEBVTT header are removed too.
Optionally, speaker labels at the start of a line (such as “Host:”) and sound tags like [Music], [Applause] and [Laughter] are removed.
Short caption lines are joined and regrouped into paragraphs of about five sentences, which makes the transcript much easier to read and summarise.
Times mentioned inside sentences, such as “at 10:30”, look the same as timestamps and may also be removed, so turn timestamp removal off if your transcript depends on them.
Worked example
A copied YouTube transcript
- Input: “0:00 welcome back to the channel / 0:03 today we’re looking at / 0:05 three ways to save on groceries”.
- Output: “welcome back to the channel today we’re looking at three ways to save on groceries”.
- A 20-minute video typically produces around 3,000 words, about a 13-minute read.
Questions people ask
How do I get the transcript of a YouTube video?
On a computer, open the video, click “…more” under the title, then choose “Show transcript”. Click the three dots in the transcript panel to turn timestamps off, or copy the whole panel and paste it here.
Can I use a cleaned transcript to summarise a video with AI?
Yes. Clean transcripts use fewer tokens and are easier for AI tools to read. Check the video creator’s terms before republishing a transcript.
Does it work with Zoom or Teams transcripts?
Yes. VTT and SRT files from meeting tools are supported. Turn on speaker-label removal only if you don’t need to know who said what.
Is my transcript uploaded anywhere?
No. The cleaning happens entirely in your browser.
Last reviewed October 1, 2026