YouTube Tools·4 min read

YouTube URL Formats Explained: Watch, Shorts, Embeds

Several different links can point at the same video, and if you write code or manage content, sooner or later you have to treat them as one. This guide looks at YouTube links from the parsing side: what each format looks like, where the 11-character video ID sits inside it, and which extra parameters come along for the ride. Once you can read the anatomy, pulling out IDs stops being guesswork and becomes simple pattern matching.

Open the free YouTube Video ID Finder All insights

YouTube URL formats, one by one

The full list.

Every one of these forms points at the same video. The ID is always 11 characters drawn from letters, digits, hyphens and underscores. Everything around it is packaging.

Watch: youtube.com/watch?v=VIDEOID, with the ID in the v parameter. Mobile m.youtube.com and music.youtube.com links work the same way
Short link: youtu.be/VIDEOID, where the ID is the first part of the path
Shorts: youtube.com/shorts/VIDEOID, same ID, vertical player
Embed: youtube.com/embed/VIDEOID, the iframe form
Live: youtube.com/live/VIDEOID, used for streams and their replays
Bare ID: the 11 characters on their own, common in APIs and spreadsheets

Where messy YouTube links come from

Share buttons are the usual source.

Share buttons add an si= token. Mobile apps add feature=shared. Playlists attach list and index parameters that bring the whole queue along. By the time a link reaches you through a support ticket or a content spreadsheet, it can carry four parameters you do not want and one you do.

A practical case: you are swapping 40 embedded videos on a documentation site, and the source list is a mix of Shorts links, timestamped shares and playlist URLs. Open the Bulk tab in YouTube Video ID Finder, paste up to 500 links one per line, and use Copy IDs, Copy CSV or Download CSV. Lines with no ID are flagged so you can fix them.

Automation makes the mess worse. A script taking links from a form or a chat export will meet every format on this page within a week, and the failure is silent: the pattern misses, the field stays empty, and nobody notices until a report ships with gaps. Clean the link at the point of entry and the rest of the pipeline stays simple.

Parsing one real YouTube link end to end

Short link in, several values out.

Paste https://youtu.be/dQw4w9WgXcQ?t=42 and the results appear as you type: Video ID dQw4w9WgXcQ, start time 42 seconds and 0:42, a Clean link of https://www.youtube.com/watch?v=dQw4w9WgXcQ&t=42s with tracking removed, a Short URL, and an Embed URL of https://www.youtube.com/embed/dQw4w9WgXcQ?start=42. A table below lists every query parameter with a short explanation.

The ID never changes across forms. Parsing is really just finding those 11 characters, discarding the wrapper, and rebuilding whichever URL shape the destination needs. Note that the tool reads the link's text only, so the ID of a deleted or private video still comes out.

Shorts and live links, the newer formats

Path segments, not parameters.

Shorts and live links trip up older parsers written when /watch and youtu.be were the only formats. Both carry the ID in the path rather than a query parameter, so a parser that only reads v= comes back empty on perfectly valid links.

The content behind them is ordinary video. A Short's ID works fine in a standard embed URL or a thumbnail request. The vertical presentation is player behaviour, not a different kind of ID.

Live links have one extra detail worth knowing: a stream and its replay keep the same ID, so a live link parsed today still works after the broadcast ends and the recording takes its place.

Timestamps in links, from t to start

Several spellings, one meaning.

Timestamps come in more spellings than seems reasonable: t=42 on shares, t=42s or t=1m30s in some links, and start=90 inside embed URLs. Share dialogs write t, while iframes only understand start. Converting between them is exactly the kind of fiddly work a parser should do for you.

The tool shows both the raw seconds and a clock value, which helps when you are writing show notes or chapter lists and need readable times next to machine-readable ones. Times over an hour show as hours, minutes and seconds, which reads far better in a description than a raw 4500. Going the other way is just as quick: type 1:30 into Timestamp link and you get watch, short and embed links that start there. For a whole list of moments, YouTube Timestamp Link Generator builds every link and a chapter list at once.

Video ID extraction mistakes worth avoiding

Where homemade parsing goes wrong.

If you have ever written your own extraction pattern, at least one of these has caught you out.

Matching any 11-character string in the URL, which happily pulls junk out of unrelated paths
Leaving underscores and hyphens out of the allowed characters, which rejects perfectly valid videos
Assuming v is the first parameter. Mobile share links often start with feature= and put v= later
Passing a t value into an embed without renaming it to start, so the timestamp never works
Assuming youtu.be links never carry query strings. They routinely carry both t and si

Tips and the neighbouring YouTube tools

IDs are the currency.

When you only need one field, copy just that row. For your own code, the regex section at the bottom of the tool has copy-ready JavaScript and Python snippets that match watch, youtu.be, Shorts, live, embed and nocookie links. Check that the result is exactly 11 characters before storing it, and keep the original URL if you might need the timestamp or playlist later.

Clean IDs feed the neighbouring tools directly. YouTube Embed Code Generator turns an ID into iframe code, YouTube Thumbnail Downloader fetches its artwork, and YouTube Channel ID Finder answers the channel question that a playlist-only or channel link cannot.

Try it now

Open YouTube Video ID Finder

The tool is one click away. No sign up, no upload, no payment.

Open YouTube Video ID Finder

Write a review