AI text detector that shows its work
Paste any text to see the signals common in chatbot writing, each one explained. Clues, not a verdict.
Free to start · No card needed · Files never leave your device
Checking a photo or video? Use the AI Detector
Signals, not a verdict
Most detectors give one number and no reasons. This one lists every signal it measured, highlights the phrases it matched and shows how much each one counts, so you can judge for yourself.
Read this before you rely on any detector
No tool can prove who wrote a text. Detectors produce false positives, and studies show they flag people who write in a second language far more often. Use the result as a starting point for a conversation, never as proof.
How it works
- 1
Paste
Paste an essay, email, post or answer you want to look at.
- 2
Review signals
See stock phrases, hidden characters, Markdown, sentence rhythm and list use, each explained.
- 3
Judge for yourself
Treat the result as a set of clues, not proof. Clean hidden characters in one click.
Private by design
Cleaning happens inside your browser. No uploads, no storage, nothing to leak.
Instant
No queue, no waiting on a server. Results show as soon as you paste or drop.
Shows its work
Every hidden character and metadata field is listed so you know exactly what changed.
Works everywhere
Clean text pastes neatly into Word, Google Docs, WordPress, Gmail and code editors.
Simple pricing
Start free. Upgrade when you need volume.
Free
$0
- ✓ Unlimited text cleaning
- ✓ Up to 5,000 characters per clean
- ✓ 5 file cleans per day
- ✓ Files up to 30 MB
- ✓ AI Detector for files and text
Pro
$7/month
or $49/year (save 42%)
- ✓ Everything in Free, no ads
- ✓ Unlimited document length
- ✓ Unlimited file cleans
- ✓ Batch: clean whole folders at once
- ✓ No file size cap
- ✓ Priority support
Questions
Can this prove a text was written by ChatGPT?+
No, and no tool can. It counts writing habits that chatbots use a lot, such as stock phrases, very even sentence lengths and heavy lists. People write that way too, and AI text that has been edited may show none of them. Treat the result as a set of clues.
Why do AI detectors flag non-native English writers?+
Detectors reward varied, less predictable wording. People writing in a second language often use simpler, more predictable phrasing, so they get flagged more. A 2023 Stanford study found popular detectors labelled most essays by non-native writers as AI-written.
What are the hidden characters it finds?+
Invisible Unicode characters such as zero-width spaces and non-breaking spaces. They often come along when text is copied out of a chat app or web page. They don't prove AI wrote the text, but they can break formatting, and the Text Cleaner removes them.
Is my text uploaded?+
No. The check runs in your browser. Nothing you paste is sent to us or stored.
Will cleaning my text change the result?+
Cleaning removes hidden characters and Markdown, so those two signals go away. The phrasing and sentence signals stay, because they are about the writing itself. We don't offer to disguise AI writing.