Skip to content
Slop or Not
Theme

Claude writing checks

Claude AI Detector: Check If Text Was Written by Claude

Paste the passage and get a verdict in seconds. Slop or Not scores the writing with a fine-tuned transformer trained on the patterns large language models fall into, so the same check that catches GPT and Gemini catches Claude. It reads the writing itself, which means it needs no watermark, no key, and no cooperation from Anthropic. Online checks run on Numen's private Mac server, away from third-party AI providers, and are deleted after processing; the iPhone and Mac apps run the check entirely on-device.

Claude Watermark: How It Works and Why It Falls Short

Catches 99.5% of AI text at a 5% false-positive rate, scored by RAID, the public benchmark, not by us.

Paste a paragraph or a whole thesis. The iPhone and Mac apps have no word cap and no monthly word budget.

Download on the App Store
  • Paste English textSchool essays, product reviews, captions, articles, and social posts up to 20,000 characters.
  • ResultA probability verdict, AI score, and readability analysis.
  • PrivacyDeleted the moment the check finishes. We never keep your text.
English text only. Deleted after processing.0/20000

AI detection is probabilistic. Results can vary with new AI models and evasion methods. Human writing that sounds formal can be flagged as AI.

Scored by a fine-tuned transformer model, not a word-pattern checker.

What it checks

The Verdict Is in the Writing Itself

Language patterns common in Claude, ChatGPT, Gemini, and other LLM writing

Sentence and paragraph regularity that reads unlike a human first draft

Optional Text Cleanup for hidden characters left behind by a copy out of a chat window

Good Uses

  • Checking an essay or assignment a student may have written with Claude
  • Screening a job application, product review, or support message that reads generated
  • Checking writing from any model, not only the ones that mark their output

Limits to Know

  • Slop or Not does not read Anthropic's text watermark, and no public tool does yet
  • Text detection is English only, so a Claude passage in another language cannot be checked here
  • Short or heavily edited writing leaves less evidence, and the verdict is a probability

FAQ

Claude AI Detector Questions

Does Claude watermark its text?

Anthropic says it does. Its August 2026 support article describes an imperceptible watermark embedded in text from models launched on or after August 2, 2026, with earlier models being retrofitted. The mark survives copy and paste and light editing, and heavy editing loses it. Anthropic also attaches C2PA Content Credentials to images Claude generates.

How can you tell if text is written by Claude?

Paste it into the checker above. Slop or Not scores the writing itself for the patterns LLMs fall into and returns a probability in seconds. The writing is the evidence that is actually available today: Anthropic has not released a tool that reads its watermark, so no one outside Anthropic can check for the mark.

Can I check Claude images here?

Not yet. The checker above is for writing. For an image, use the C2PA Content Credentials checker linked below: image files Claude generates carry signed C2PA provenance, and Slop or Not reads those credentials and falls back to pixel-level detection when a file carries nothing.

Can you remove a Claude watermark?

Anthropic says heavy editing, rewriting, or converting the file can lose the mark. Text Cleanup in Slop or Not removes mechanical fingerprints such as invisible characters and smart punctuation. It does not touch a statistical watermark, and we make no claim that it does.

Does the absence of a watermark prove text is human?

No, and Anthropic concedes the same point: a missing mark proves nothing. Most models leave no mark at all, and heavy editing removes the ones that do. That is why Slop or Not reads the writing, and why the verdict is evidence for a person to weigh rather than a ruling.

When should I use the AI text detector instead?

Both pages run the same check. Use this one when Claude is the suspected source and the watermark question is what brought you here. Use the AI text detector for essays, resumes, and messages when the model behind them is unknown.

Run the Next Check on Your iPhone or Mac

The Slop or Not app checks images and text on-device, so nothing you check ever leaves your device.

Download on the App Store

Related checks

Compare Related Detector Workflows

AI Text Detector

Check whether writing looks AI-generated with Slop or Not. Paste text online for a quick check, or use the iPhone and Mac apps on-device.

AI Watermark Detector

Use this AI watermark detector on images for SynthID, C2PA, and IPTC evidence, or check writing with the RAID-scored text detector.

C2PA Detector

Use this C2PA detector to check Content Credentials with Slop or Not. It reads signed provenance, then checks pixels when credentials are absent.