AI slop detector: is it slop or not?
Paste a post or an x.com link. Jev, a decision model, says whether it reads as AI slop, and every line that gives it away is marked in place, with a plain swap wherever one exists.
- 100% free
- No login
- Works with x.com links
- 20 checks a day
Verdict
Readers will likely take it for AI writing.
- Generic, no first-hand detail
- 91%
- No voice of its own
- 85%
- Written to farm engagement
- 81%
๐ In today's fast-paced digital landscape, it's not just about posting more, it's about posting smarter. Let's dive in: here are 3 game-changing strategies to unlock your growth and elevate your personal brand. ๐กโจ Furthermore, it's important to note that consistency is paramount. What do you think? Let me know in the comments! #growth #buildinpublic
What gives it away
- Formula
It's not just X, it's Y
The favourite AI construction. Say the second half on its own.
- Opener
In today's world
A warm-up that says nothing. Start with the point.
Swap โIn today's fast-paced digital landscape,โ for โToday,โ
- Opener
Stock opener
A stock hook. Open with your sharpest line instead.
- Stock phrase
Game-changer
Say what it changed instead.
- Stock phrase
Hype verb, 2 times
Promises a transformation without saying what changes.
- Bait
Comment bait, 2 times
Asks for engagement instead of earning it.
- Emoji
Emoji chain
Stacked emoji read as decoration, not voice.
- Filler
Furthermore
Formal glue nobody uses out loud.
Swap โFurthermoreโ for โAlsoโ
- Filler
It's important to note
If it matters, just say it.
Cut โit's important to note thatโ
- Stock phrase
Buzzword
Sounds impressive, carries no information.
- Hashtag
Hashtag, 2 times
Hashtags read as marketing and do not help reach on X.
- Emoji
Stock emoji
The rocket and sparkles are the first emoji AI reaches for.
Over 240 characters: the feed clips it.
How to check if a post is AI slop
- 1
Paste a post or a link
Type or paste a post, a thread or a LinkedIn draft, or the link to any public post on X. It opens on an example so you can see how it works first.
- 2
Get Jev's verdict
Press Check it. Jev, a decision model, reads the whole post and answers one question: does it read like an AI model wrote it? You get slop, borderline or not slop, how sure it is, and why.
- 3
See every tell
Stock phrases, formulas like it's not just X, it's Y, filler, stacked emoji and comment bait are marked in place as you type, heaviest first.
- 4
Fix it and check again
One click makes the safe swaps, such as utilize to use and furthermore to also. Rewrite the formulas yourself, check again, and copy it once it reads like you.
How it decides
Two checks, each doing what it is good at. The phrase rules run in your browser as you type and mark every tell they know, with a swap wherever a safe one exists. Jev reads the whole post and answers one question: does it read like an AI model wrote it?
They can disagree, and that is useful. A post with no stock phrases can still read as generic, which Jev catches; a post Jev passes can still carry a phrase worth swapping. The verdict is Jev's: slop at 60% or more, not slop under 40%, borderline in between.
The tells it looks for
- Stock phrases: delve, tapestry, a testament to, navigate the complexities, leverage, game-changer, seamless, robust.
- Hype verbs that promise a change without naming it: unlock, unleash, elevate, supercharge, revolutionize.
- Formulas: it's not just X, it's Y; not only, but also; No X. No Y. Just Z.
- Warm-up openers and essay filler: in today's fast-paced world, let's dive in, it's important to note, in conclusion.
- Habits: em dashes, stacked emoji, checkmark bullets, hashtags, and endings that ask for comments.
Frequently asked questions
What is AI slop?
Writing that reads as machine-made: stock phrases like delve and game-changer, formulas like it's not just X, it's Y, filler such as it's important to note, stacked emoji and comment bait. Readers have learned to spot it and scroll past.
Can this tell if a post was written by AI?
No tool can prove who wrote a post. Jev judges how it reads: whether readers will take it for AI writing. A person can write slop, and a carefully edited AI draft can come out clean.
What is Jev?
Jev is a decision model from TypeSafe AI. It does not write anything: it reads a post and answers yes-or-no questions with a probability. Here it answers whether the post reads like an AI model wrote it, plus three questions behind that: is it generic, does it have a voice of its own, and is it written to farm engagement.
Can I check someone else's post?
Yes. Paste the link to any public post on X and the tool reads the post, long posts included, then judges it. Protected and deleted posts cannot be read.
Why are em dashes flagged?
Heavy em dash use has become the most talked-about sign of AI writing, fairly or not. One is fine; a post full of them reads as generated. A comma or a full stop usually works better on X.
How do I make a post sound less like AI?
Lead with the point, use the words you would say out loud, add a detail only you know, cut the filler and the emoji, and end on something specific instead of asking for comments. Xpert drafts from your own past posts, so its drafts start in your voice.
Does the AI slop detector store my text?
No. The tells are marked in your browser as you type. When you press Check it, the text or link goes to Xpert's server and on to Jev (TypeSafe AI's model, reached through OpenRouter) for the verdict. Typed text is not saved; a public post checked by its link is kept for a few hours, so the next person to check it gets the answer at once. You get 20 checks a day with no account.
More free tools
- Viral Post CheckerWill it go viral? A 0 to 100 score, the odds, and every reason with a fix, in under a second.
- Tweet GraderScore your draft 0 to 100 before you publish, with a rewrite that fixes it.
- AI Tweet ImproverPaste a draft and get 3 stronger rewrites that keep your idea and facts.
- Character CounterCount characters and words and stay inside the 280 limit.
Want more than a one-off tool?
Xpert writes in your voice and keeps score.