A YouTube transcript
with no clock.

01 / THE FILE WITH NO CLOCK

The version without timestamps is the TXT file.

Three downloads come out of the YouTube transcript downloader, and only one of them carries no time codes at all. The format table on that page says it plainly: .txt is the text in order, without timestamps. .srt is numbered captions with a start and an end time on each cue. .vtt is the same timed cues wrapped for a web player.

So the shortest route to a plain transcript is to leave the format menu on TXT and download. No conversion, no second tool, no editing the file afterwards to strip numbers out of the front of every line.

It is worth knowing what the plain text actually is, because it is not a rewrite. The words are the caption text in the order the captions carry them, with the timings dropped. For reading and searching, that is exactly what you want. If you are pasting into a document with its own layout, expect to reflow the paragraphs yourself.

02 / THE CHECKBOX

The timestamps checkbox changes the view, not the file.

The result panel has a checkbox labelled Show timestamps, and it starts ticked. Unticking it removes the time column from the transcript on screen, which is what most people want when they are reading rather than following along with the video. It does not change what the download buttons write.

That is worth being explicit about, because the expectation runs the other way: a checkbox next to a download menu looks like a setting for the file. It is a display toggle for the reader on the page. The file that leaves the page is decided by the format menu alone, and TXT carries no clock whether the box is ticked or not.

The inverse is equally true. Ticking the box does not add timings to a TXT download, and unticking it does not strip them out of an SRT. If you want timed cues, choose SRT or VTT; if you want clean prose, choose TXT. The checkbox is only about how the transcript looks while you read it.

03 / GETTING THE TEXT AT ALL

First, the transcript has to exist.

None of this works until the page has the text, and on this route the text comes from a track that was attached to the video beforehand, whether the creator wrote it or YouTube generated it. Paste a watch link, a youtu.be link, a Shorts, live or embed link, or the eleven-character video ID, and the tool asks YouTube for those tracks and shows the one it picked.

A video with no caption track has nothing to download here. That is not a limit anyone can work around on this page, because the link route does not listen to the soundtrack at all. The message says so, and the page offers the alternative: if it is your own video, upload the original file and speech recognition runs on the audio.

Two related limits are worth knowing about, both covered in more detail in the walkthrough on how to get a transcript of a YouTube video. Whole playlists and channels are not offered, and neither are translated captions: one video at a time, in the languages that video already carries.

04 / WHEN YOU WANT THE TIMES

When you do want the times.

The timings are not invented for the download. Each caption event arrives with a start time and a duration, and the SRT file writes the start as hours, minutes, seconds and milliseconds with a comma before the milliseconds, and the end as the start plus the duration. VTT keeps the same cues with a full stop instead of a comma and a WEBVTT line at the top.

If you are unsure which one a program wants: SRT is the older of the two, and the one subtitle editors tend to expect. VTT is the one a web player uses. Both are built in your browser from the transcript on screen, along with the TXT.

There is one honest caveat about timed files and it comes from the source, not from the download: caption events can overlap, because on-screen captions do. The download keeps that overlap rather than tidying it, so a cue can end after the next one has started.

05 / WHEN THE LINK FAILS

When the link route has nothing to give.

Two failures are worth planning for.

The video has no captions. No caption track means no transcript from a link here, and the page says so rather than returning an empty result. For your own recording, upload the file instead: the same page accepts an audio or video file and runs speech recognition on it, then offers the same TXT, SRT and VTT downloads. That is the route that works without any captions at all.

YouTube refuses the request. When that happens, the page offers a second way: open the video on YouTube, use the transcript panel there, copy the text, and paste it into the box on the page. It is tidied in your browser and downloads exactly like a transcript fetched by the tool, so you can still end up with a clean TXT file. On a phone, the YouTube app shows a transcript but does not let you copy it, so this route is a computer job.

One consequence is easy to miss. A pasted transcript is somebody else's text, so any spelling it carries is the spelling you keep. Tidying works on the structure, not the words: it reads the pasted lines, drops the time stamps and the labels around them, joins the lines that belong to one cue, and writes the plain text the download buttons use.

06 / QUESTIONS