Paste your title and description and get a ranked list of tag ideas — single keywords and two- or three-word phrases, ordered by how often they appear, with filler words skipped. The result is a comma-separated list you can paste straight into the tag field, sized to fit YouTube's budget. That budget is the part most tools get wrong: the field holds about 500 characters in total, and a tag containing a space is stored quoted, so a two-word tag costs two characters more than it looks. The meter here charges them. One thing to be clear about: this works on text you paste. Reading another video's hidden tags requires the YouTube Data API and a server, so no browser tool can do it.
| Type | Example | Budget cost |
|---|---|---|
| Single keyword | tools | 5 characters |
| Two-word phrase | browser tools | 15 — quotes are charged |
| Separator | comma between tags | 1 each |
| Total budget | ~500 characters | Across all tags |
No, and be sceptical of anything that claims to do it in the browser. Hidden tags come from the YouTube Data API, which needs a server and an API key. This tool analyses the text you give it.
Because that is roughly what YouTube's tag field holds in total across all tags. Go over and the rest are silently dropped on save. The meter also charges the two quote characters a multi-word tag costs, which is why a list that looks like 498 characters elsewhere may not fit.
By how often each keyword or phrase appears in your text. When two tie, the longer phrase wins, because "browser tools" is a more specific tag than "tools".
Because phrases are only built inside a run of words uninterrupted by punctuation or a line break. Most n-gram extractors skip that and produce things like "subscribe chapters" from two unrelated lines.
Less than they once did — YouTube has said they play a minimal role for discovery, mostly helping with common misspellings of your topic. Your title, description and thumbnail do far more. Fill the field, but do not agonise over it.
Yes. Set the maximum phrase length from one to four words, and a minimum frequency so one-off words drop out.
No. The counting happens in the page.
The text is lowercased and split on anything that is not a letter, digit or an internal apostrophe or hyphen — which includes newlines, emoji and list bullets. That split is the part that matters, because phrases are only built inside a run of words uninterrupted by punctuation or a line break. Without it, a description ending one line with "subscribe" and beginning the next with "chapters" yields the phrase "subscribe chapters", which nobody wrote and which is useless as a tag; most simple n-gram extractors produce a list full of exactly that. Unigrams and phrases are counted together, with a phrase discarded if it begins or ends on a filler word, so "the best" and "tools for" never appear while "best browser tools" can. Ranking is by frequency, with a longer phrase winning a tie against a shorter one on the reasoning that "browser tools" is a more useful tag than "tools" when both occur the same number of times. The stopword list is the one the summariser and the resume scanner already use, rather than a second copy to drift out of sync. The budget meter is where this differs from a plain keyword counter: YouTube's tag field holds roughly 500 characters in total, not per tag, and a tag containing a space is stored quoted — so a two-word tag costs two characters more than it looks. Counting without the quotes is how a list that measured 498 characters gets silently truncated on save, so the meter charges them, and charges the separating comma between tags but not one before the first. Tags are then taken in rank order until the budget is spent, with anything that still fits kept even if a longer tag ahead of it was skipped. One honest limitation stated on the page: this reads the text you paste. Another video's hidden tags are only available through the YouTube Data API, which needs a server and a key, so no browser tool can show them.