FREE CHECK

AI Slop Checker and Detector

Updated November 16, 2026

Paste a draft into this free AI slop checker. You get a slop score out of 100 and a list of each AI tell, with the line it sits on.

AI slop is writing that shows a model's habits, such as stock words and an ending that repeats the piece. The checker scans your draft against 23 documented patterns and gives it a score. Then it hands back an edit with those patterns removed. It never guesses who wrote the text.

Best for newsletter writers and bloggers who draft with AI and want the piece to read like them before it goes out.

Free once a day on one model, or three times a day with a free account, shared with the AI Humanizer. Council runs four models on any paid plan, from $20 a month.

Depth

Council comes with any paid plan, from $20 a month. No Deep on this check: a debate doesn't improve an edit.

0 / 2,000

Want the edit more than the score? The AI Humanizer runs the same check

What counts as an AI tell?

An AI tell is a word, a sentence shape or a habit that Wikipedia's editors list as a sign of AI writing. This checker looks for each one in four steps.

1

We scan your draft against 23 documented patterns

Stock words such as "delve", "leverage", "robust" and "tapestry". Dead phrases such as "it is worth noting". Habits such as lists of three, slow openers and endings that repeat the piece. You get a score out of 100 before the edit starts.

2

We check how the prose reads too

This step looks for passive voice, adverbs that prop up weak verbs, long sentences and a high reading grade. It gets its own score, apart from the AI tells.

3

We edit, then run both checks again

The editor gets the exact list of faults. Then the same scanner runs on the edit, so you can read both scores side by side.

4

We point at what an edit cannot fix

The editor quotes each place where your draft makes a general claim, hedges a number or cites nobody. It says what only you can add: an example you lived, a figure from your own data or a named source with a date. It writes none of it for you.

The pattern list comes from Wikipedia's Signs of AI Writing guide, kept by WikiProject AI Cleanup. We added nine rules for how prose reads. Both run as fixed rules, so the same text always gets the same score.

A draft with no AI tells can still hold a made-up fact. TrueStandard checks every claim in a draft across four models in about 60 seconds. Check Your Draft →

On 26 September 2026 we ran these checks on 629 drafts from six AI models: Claude Sonnet 5 and Haiku 4.5, GPT-5.6 Terra and Luna, and Gemini 3.7 and 3.8 Flash. Claude used the most em dashes, GPT the most lists of three and Gemini the most stock AI words. Read the full test: which AI uses the most em dashes. On 20 November 2026 we ran them on 1,155 drafts from 11 models, grouped by size: the best AI model for writing.

Our guide covers what AI slop is and how to avoid it.

Before and after

We wrote this sample draft to show the check. Both counts come from running the scanner on this exact text.

Draft as pasted AI tells: 12

In today's world, the landscape of creator monetization is undergoing a transformative shift. It's worth noting that many creators are now leveraging multiple revenue streams to foster sustainable growth. Here's the thing. Most creators delve into paid tiers far too early. The reality is that a seamless upgrade path takes 6 to 12 months to build properly. And that matters.

After editing AI tells: 0

Creator monetization is changing. Many creators now run several revenue streams at once to keep growing. Most of them open paid tiers too early, before the free list can carry a paid offer. Building an upgrade path takes 6 to 12 months.

Frequently Asked Questions

Powered by TrueStandard

Four independent models check every claim in your draft.