How to Find a YouTube Video ID in 2026 (Every URL Format)

Published 1 September 2026 · 12 min read

Every video on YouTube has one permanent name, and it is not the title. It is an eleven-character code called the video ID — the thing that sits after v= in a watch link, after the slash in a youtu.be share link, and after /shorts/ in a vertical video URL. Titles change, thumbnails get swapped, channels get renamed. The ID never moves. Once you can read it out of any link on sight, a surprising amount of YouTube work gets easier: embedding, thumbnail fetching, API calls, competitor tracking, and every spreadsheet that needs to say “this exact video” without ambiguity.

This guide covers what the ID actually is, how to find it in all nine URL shapes YouTube uses, how to pull it out in bulk, the mistakes that silently break embeds, and the practical things you can build once you have it. If you just want the code right now, paste any link into the YouTube Video ID Finder and it will strip the URL down for you.

What a YouTube video ID actually is

A YouTube video ID is a fixed-length string of exactly 11 characters, drawn from a 64-character alphabet: the 26 uppercase letters, the 26 lowercase letters, the digits 0–9, plus the hyphen (-) and the underscore (_). That is base64url encoding, and it yields roughly 73 quintillion possible combinations — enough that YouTube has never needed to lengthen the format despite indexing billions of uploads.

Three properties matter in practice:

• It is permanent. The ID is assigned at upload and survives every later edit. A video can change title five times and keep the same ID.
• It is case-sensitive. An uppercase and lowercase version of the same letters are two different IDs. This breaks more embeds than any other single mistake.
• It is universal. The same 11 characters work in a watch URL, an embed, a Shorts link, an API request and a thumbnail CDN path. One key, every surface.

It is worth separating the video ID from the other IDs YouTube hands out, because they are easy to confuse. A channel ID is 24 characters and starts with UC. A playlist ID starts with PL, UU, LL or similar and appears in a list= parameter. A handle looks like @channelname. None of those are video IDs, and none of them can be converted into one — you have to open an actual video. If it is the channel identifier you need instead, the Channel ID Finder handles handles, custom URLs and legacy /c/ links.

Every URL format, and where the ID hides in it

YouTube has accumulated a lot of link shapes over twenty years. Here is where the ID sits in each one. In every example below, dQw4w9WgXcQ stands in for the ID.

• Standard watch page: youtube.com/watch?v=dQw4w9WgXcQ — after v=.
• Short share link: youtu.be/dQw4w9WgXcQ — the entire path.
• Shorts: youtube.com/shorts/dQw4w9WgXcQ — after /shorts/.
• Embed: youtube.com/embed/dQw4w9WgXcQ — after /embed/.
• Live stream: youtube.com/live/dQw4w9WgXcQ — after /live/.
• Mobile: m.youtube.com/watch?v=dQw4w9WgXcQ — identical to desktop, different subdomain.
• Privacy-enhanced embed: youtube-nocookie.com/embed/dQw4w9WgXcQ — after /embed/.
• Legacy inline: youtube.com/v/dQw4w9WgXcQ — after /v/, still found in old CMS content.
• Attribution link: the ID appears inside a URL-encoded u= parameter; decode it first, then read v=.

The pattern is consistent: the ID is either the value of the v= query parameter or the last path segment. Everything else in the URL is optional decoration.

The parameters that are not part of the ID

The most common extraction error is copying too much. YouTube appends tracking and playback parameters after the ID, and they look enough like part of the code to survive a careless copy-paste:

• ?si= — the share-session token added by the Share button. Long, random, and not part of the ID.
• &t= or ?t= — a timestamp in seconds. Harmless in a browser, fatal in an API call.
• &list= and &index= — playlist context, added when the video was opened from a playlist.
• &feature=, &pp=, &ab_channel= — attribution and experiment flags.

The rule is simple: the ID stops at the first & or ? that follows it, and it is always exactly 11 characters. If your string is 12 or more, you have taken a passenger along. If it is 10, you truncated it. Counting the characters is the fastest sanity check there is.

Finding it on mobile and in the app

The mobile app hides the address bar, so the workflow differs slightly. Tap Share under the video, then Copy link. You will get a youtu.be link with an ?si= token attached — the ID is the segment between the final slash and the question mark. On Shorts the Share sheet returns a /shorts/ URL with the same structure.

One useful trick: any Shorts ID also works as a regular watch link. Replace youtube.com/shorts/ID with youtube.com/watch?v=ID and the same video opens in the standard player, where the full description, the channel context and the desktop controls are all visible. That is often the quickest way to research a Short properly rather than scrolling past it.

Extracting IDs in bulk

If you are auditing a niche you will not be doing this one link at a time. Three approaches, in increasing order of effort:

• Paste-and-strip. Drop each URL into the Video ID Finder and copy the clean ID. Fine for a dozen videos, tedious past that.
• Spreadsheet formula. In Google Sheets, a single formula handles the two dominant formats: take the text after v= when it exists, otherwise the last path segment, then trim to 11 characters. Paste a column of URLs, fill down, and you have a clean ID column you can sort and deduplicate.
• Regular expression. For developers, one pattern covers nearly everything: match v=, /shorts/, /embed/, /live/, /v/ or a youtu.be/ host, then capture exactly 11 characters from the allowed alphabet. Anchor the capture to that length and the tracking parameters fall away on their own.

Whichever route you take, deduplicate on the ID rather than the URL. The same video reaches you as a watch link, a share link and an embed, and only the ID reveals that they are one video.

What you can actually do with the ID

The ID is not trivia — it is the input to most of the useful things you can do outside the YouTube interface.

• Fetch thumbnails directly. YouTube's image CDN builds paths from the ID, which is why a thumbnail can be pulled without an API key. The four available sizes and their fallback order are covered in the thumbnail download guide, and the Thumbnail Downloader does it in one paste.
• Build clean embeds. An /embed/ID iframe with no playlist or autoplay junk loads faster and behaves predictably across browsers. Use the nocookie domain when you care about consent banners.
• Read competitor metadata. Tags are not visible in the interface, but they are in the page source keyed to the ID — the tags extractor surfaces them, and the tag strategy guide explains what to do with them.
• Check availability by country. Region blocks are applied per video, so the ID is what you feed the region checker to see where a video is watchable.
• Make timestamped links. Appending &t=90 to a watch URL starts playback at 90 seconds — useful for citing a specific claim in a long video.
• Track research over time. Log competitor uploads by ID in a sheet with title, publish date and view count. Because IDs are permanent, you can re-poll the same rows for months and see which titles and thumbnails were quietly changed.

Common mistakes that break things

• Lowercasing the ID. Some CMS fields and URL sanitisers normalise case. The video vanishes and the embed shows “Video unavailable” with no explanation.
• Keeping the si= token. It works in a browser because YouTube ignores extras, but any API request built from that string fails validation.
• Confusing channel and video IDs. A UC… string in a video field returns nothing. Check the prefix and the length before debugging further.
• Retyping from a screenshot. l versus I, 0 versus O, hyphen versus underscore. Always copy, never transcribe.
• Assuming the ID implies availability. A valid ID can point at a private, deleted or region-blocked video. If a link fails, the video-not-available guide walks through the causes.

The takeaway

The video ID is the smallest useful unit of YouTube. Eleven case-sensitive characters, identical across every URL format the platform has ever shipped, permanent for the life of the upload. Learn to spot it after v=, after a youtu.be/ slash and after /shorts/, stop reading at the first ampersand, and confirm the length is exactly 11. From there it feeds thumbnails, embeds, tag research, region checks and any tracking system you build — which is why the least glamorous string on the platform is also the most useful one to know by heart.