AI slop detector
Find the slop before your readers do
AI slop is text that fills space without saying anything. The AI slop detector reads every sentence, flags the filler, hype, hedges and recaps, and names the reason for each. It is free and never rewrites a word.
Free on every plan. It never uses up a page.
Made for
- Blog posts and newsletters before they publish
- Work from freelancers, agencies or a content team
- Pages you suspect were mass-produced
What it flags
Example: Slop 83 · 5 of 6 sentences flagged
In the ever-evolving world of remote work, productivity has never been more important.Slop: filler But what does it really mean to be productive?Slop: set-piece rhetoric In our survey of 1,600 employees, hybrid workers were 33% less likely to quit than colleagues in the office full time. This groundbreaking finding is a game-changer for companies everywhere.Slop: hype It could potentially suggest that flexibility might, in some cases, matter.Slop: hedging In short, remote work is here to stay.Slop: recap
How Slop check works
Step 01
Every sentence is read
Each sentence is judged on what it does in its passage, not on which words it uses, so it works in every language.
Step 02
Padding is named
Filler, inflated wording, hedging, set-piece rhetoric, recap or hype: each flagged sentence carries its reason.
Step 03
You get a score
The share of sentences that are padding, from 0 (none) to 100 (all of it).
Step 04
Fix it in one click
Run De-slop on the same text to cut what was flagged, with every fact checked.
Options
- Free
- Slop checks never use up a document, on any plan.
- Any source
- Paste text, enter a web address or upload a PDF.
From code, terminal or agent
# REST API
curl https://readable.md/api/v1/transform \
-H "Authorization: Bearer $READABLE_API_KEY" \
-H "Content-Type: application/json" \
-d '{ "tool": "slop-check", "text": "…" }'
# Command line
pbpaste | readable check
# MCP tool: slop_checkQuestions
- What counts as AI slop?
- Text that takes up room without adding anything: stock openers, inflated wording, a hedge on every claim, rhetorical set pieces, summaries of what was just said, and hype. Each flagged sentence is labeled with one of these.
- Is it an AI detector?
- No. It does not guess who wrote a text, and people write slop too. It shows which sentences are padding, which is what readers notice.
- What does the score mean?
- The share of sentences flagged as padding, from 0 (none) to 100 (all of it).
