Инструменты для соцсетей IT Market
Вставьте автосубтитры YouTube с таймкодами и получите чистый текст для статьи, описания, саммари или поста.
Tool guide
The "Show transcript" panel under a video hands you a wall of text with a timestamp before every line, and an SRT file adds cue numbers and arrow lines on top of that. Paste either one in and you get continuous prose back — ready for notes, an article, a video description or a pass through a translator. Parsing happens locally in the browser, so an internal webinar transcript can be cleaned without handing it to someone else's service. See also: remove the timestamps from an SRT file, turn the text back into SRT subtitles, turn the transcript into a video description.
Under the player expand "...more", choose "Show transcript", turn off timestamps in the panel if that toggle is offered, then select the panel contents and copy. The mobile app does not always expose transcripts, so opening the video in a desktop browser is usually faster.
SRT with cue numbers and arrow lines works reliably, as does a plain copy-paste from the transcript panel where the time sits on its own line as 1:23. WebVTT stamps like 00:00:01.000 are stripped too, though the file header and any styling lines are best deleted by hand.
The tool only removes; it never writes. YouTube's automatic captions ship without punctuation in most languages, and restoring it takes a human or a language model. Treat the output as raw material for editing rather than a finished draft.
Yes, bracketed non-speech markers go out along with the timecodes. If your transcript uses a different convention — parentheses, or the ♪ symbol — you will need to clear those with find-and-replace in a text editor.
No hard limit; the work happens in the tab's memory. An hour-long video is roughly 8,000-10,000 words and cleans instantly. On a low-end phone, pasting a very large block into the field can lag more than the cleaning itself does.
No, there is no summarisation here — only the removal of technical markup. The clean text is easy to pass onward: into a translator, into your notes, or into a language model, which handles a continuous paragraph far better than a wall of timecodes.
It does not. A script inside the page does the parsing and no request carries the field contents anywhere, which is why transcripts of private meetings and internal training recordings can safely go through it.
Before: 1 00:00:01,000 --> 00:00:03,400 today we are setting up 2 00:00:03,400 --> 00:00:06,900 a microphone for podcast recording
After: today we are setting up a microphone for podcast recording
Before: 0:14 hey everyone 0:16 [Music] 0:19 let us start with the basics
After: hey everyone let us start with the basics
Your rating and feedback help decide what to improve next.
