Yes, with no account. Fifteen checks an hour and sixty a day per visitor. It reads published numbers and runs no models, so it costs almost nothing and the limits are there to stop scraping rather than to ration it.

Why does it say a signal was not measured?

Because it could not be. Comment counts are not published on the page this reads, so the signal that needs them reports as unmeasured rather than being filled in with a guess. The score says how many of the four it is actually built on.

Does this work for TikTok or Instagram?

Not yet. It reads YouTube. The same argument applies on those platforms, and the follower-list problem is if anything worse there.

The subscriber number looks rounded.

It is, and not by us. YouTube rounds it before publishing it, so 4.52M is what the platform says rather than what the channel has. Every ratio built on it inherits that, which is fine for a comparison and wrong for anything presented as exact.

← Free tools

Free fake follower check

Paste a YouTube handle and this reads the creator's recent videos and reports whether their audience behaves like a real one: what share of subscribers actually watch, whether engagement matches the platform norm, and whether views are steady or spiky. It does not report a percentage of fake followers, because that number cannot be calculated from anything YouTube publishes. Free, no account.

The honest version of a tool everybody else fakes.

Why this does not give you a percentage

Because nobody can. YouTube does not publish who subscribes to a channel. There is no list to sample, no way to look at a thousand accounts and count how many are bots, and the subscriber figure the platform shows is rounded before you ever see it: a channel displayed as 4.52M is somewhere in a band around that, never exactly it.

So when a tool tells you a creator has 23% fake followers, ask where the number came from. On YouTube it came from a model that has never seen the subscriber list, because no such list is available. It is a number with the shape of a measurement and none of the substance.

What can be measured is behaviour. An audience that was bought does not watch, does not engage in proportion, and produces view counts that either barely move or swing wildly. Those show up in numbers YouTube does publish, and that is what this reads.

What the signals actually mean

Reach against audience size is the one that catches a purchased audience. A channel with two million subscribers where a typical video gets eleven thousand views has a problem that no amount of follower count hides. The comparison is against what is normal at that size, because big channels reach a smaller share of their subscribers than small ones do, and judging them by the same bar would flag every large creator.

Engagement rate is the weakest of the four, and it is included because its absence would be conspicuous rather than because it is good. It is easy to inflate, easy to deflate by posting about something serious, and it says more about a topic than about an audience.

View consistency catches a different failure: a channel whose views barely vary across a dozen videos is behaving less like an audience and more like a number being maintained.

What this cannot tell you

Whether any individual subscriber is real. That question needs data the platform does not release to anybody, including the tools that answer it confidently.

Whether the creator bought them. A channel can have poor reach because the audience is bought, because the topic changed, because a video went viral to people who never came back, or because they post rarely. The signals describe the audience's behaviour, not the creator's intent, and treating one as the other is how a fair creator ends up on a blocklist.

Anything about who the audience is. Age, location and interests are not visible here, and a check that guessed at them would be inventing the most commercially useful part of the answer.

Common questions

Is this free?: Yes, with no account. Fifteen checks an hour and sixty a day per visitor. It reads published numbers and runs no models, so it costs almost nothing and the limits are there to stop scraping rather than to ration it.
Why does it say a signal was not measured?: Because it could not be. Comment counts are not published on the page this reads, so the signal that needs them reports as unmeasured rather than being filled in with a guess. The score says how many of the four it is actually built on.
Does this work for TikTok or Instagram?: Not yet. It reads YouTube. The same argument applies on those platforms, and the follower-list problem is if anything worse there.
The subscriber number looks rounded.: It is, and not by us. YouTube rounds it before publishing it, so 4.52M is what the platform says rather than what the channel has. Every ratio built on it inherits that, which is fine for a comparison and wrong for anything presented as exact.

Free fake follower check

Why this does not give you a percentage

What the signals actually mean

What this cannot tell you

Common questions

A rate tells you the size of an audience. It does not tell you who the creator is on camera.