SRT vs VTT YouTube Guide: Caption Files Explained, Plus SBV

The SRT vs VTT YouTube question usually comes up when you are about to upload captions, or you have downloaded a file and want to know what is in it. The short answer is that the difference is mostly in punctuation and styling, not in the text.

This page uses Google's list of supported subtitle and closed caption files, checked in September 2026, for what YouTube accepts. For the WebVTT format itself, it uses the W3C specification.

What a caption file contains

Google puts it simply: a caption file "contains the text of what is said in the video. It also contains time codes for when each line of text should be displayed." Some formats add position and style information on top.

Every format below is a series of short blocks, often called cues. Each has a timing line with a start and an end, followed by one or two lines of text. Everything else is detail.

Here are the first two blocks of Google's own SRT example, exactly as the help page shows them. The number, the timing line and the text are the whole of it.

1
00:00:00,599 --> 00:00:04,160
>> ALICE: Hi, my name is Alice Miller and this is John Brown

2
00:00:04,160 --> 00:00:06,770
>> JOHN: and we're the owners of Miller Bakery.

Notice that the speaker names are just part of the text. None of these formats has a special field for who is talking, so a caption file only shows speakers if whoever wrote it typed them in.

SRT vs VTT YouTube files: reading the timing lines

SRT (SubRip). Each block starts with a number, then the timing line. Google's example is 00:00:00,599 --> 00:00:04,160: hours, minutes, seconds, a comma, then milliseconds, with an arrow between start and end.

VTT (WebVTT). The file must begin with the word WEBVTT. Its timing lines use the same arrow but a full stop before the milliseconds, as in 00:00:00.599 --> 00:00:04.160. The W3C specification makes the hours optional unless they are non-zero, and cue numbers are optional too.

SBV (SubViewer). No numbers and no arrow. Google's example is 0:00:00.599,0:00:04.160, with a comma between start and end. Google says the main difference between SRT and SBV "is the format of the caption start and stop times."

What YouTube supports in each

Google files SRT and SBV under basic formats. For both, it says: "Only basic versions of these files are supported. No style info (markup) is recognized. The file must be in plain UTF-8."

WebVTT sits under advanced formats, with a caveat. Google's note reads: "In initial implementation. Positioning is supported, but styling is limited to <b>, <i>, <u> since CSS class names are not yet standardized."

So a VTT file with elaborate styling will lose most of it on YouTube. If you need full control of styling and positioning, Google points to TTML, which it lists as in partial implementation with both supported.

Which one to use

For most uploads, the choice barely matters. Google's own advice for newcomers is SRT or SBV, because "they only require basic timing info, and can be edited using any plain text editing software."

• Pick SRT if you want a simple format you can edit in any text editor, with numbered blocks that are easy to find.
• Pick VTT if you need captions placed on screen, for example to avoid covering on-screen text.
• Pick SBV only if a tool you already use exports it. It offers nothing SRT lacks on YouTube.

Whichever you use, save it as plain UTF-8 text. Google names that requirement for the basic formats, and a different encoding risks garbled characters.

Timing lines are what make a file useful

Strip the timing lines out of any of these files and you have a plain transcript: the words, with no link back to the video. That is what most people end up with when they copy a transcript and switch timestamps off. The free transcript cleaner does exactly that to an SRT or VTT file in one paste.

Google treats the two differently on upload too. When adding captions in YouTube Studio, you choose between a file "With timing" or "Without timing". A file without timing has to be synced to the video before it works as captions.

The guide to copying a YouTube transcript covers what else the copy loses. The timing is the part that is hardest to get back.

One practical note on where these files come from. Google's caption editing page describes downloading a caption file for your own videos in YouTube Studio, then editing it in a plain text program such as TextEdit or Notepad and uploading it again. It does not describe a download for other people's videos.

Which parts a summarizer uses

A summarizer that reads captions needs two things from a cue: the text, and the start time. The text is what gets summarized. The start time is what lets a point in the summary link back to the moment it came from.

The rest is ignored. Cue numbers, the WEBVTT header, end times, styling tags and positioning carry no meaning for a summary. The quality of the words matters far more than the format, which is why timestamps you can check are worth more than tidy styling.

Our tool reads the caption track YouTube already has, so you never upload a file. On the free plan you get a TL;DR, key takeaways, timestamped chapters and full notes, in whatever language the video is spoken in and at any length. The chapters use the timing lines, which is what those cue times are for.

We don't export SRT or VTT

To be plain about it: our summarizer does not produce SRT, VTT or SBV files. A summary is not a caption track — it has no cue for each line of speech — so it would not make sense as one.

Export is copy to clipboard or Markdown, which pastes cleanly into notes apps and documents. If you need a caption file itself, Google describes downloading one for your own videos from YouTube Studio, under Subtitles, with Options and then Download subtitles.

Frequently asked questions