Was this text written by a person, or generated by AI?
Paste a text and ParaTrace tells you in seconds, marking the parts that look AI-written. Try it free, with no account needed.
99.8 %
of AI answers caught
1.0 %
of human texts wrongly flagged
90
AI models tested
Level with Pangram on AI-written essays, and like Pangram it flags none of 106 research papers written before ChatGPT. Compare with Pangram
Free checks aren't stored. To count them, the server keeps your network address in memory until midnight UTC. Privacy policy
By the numbers
Read the benchmark report99.8 %
of AI-written answers detected
25,601 answers · 26 developers
1.0 %
of human texts wrongly flagged
22,681 held-out texts, same setting
95.9 %
still detected under evasion attacks
11 attack types
161 ms
to score a 300-word text
on our servers
How it works
From text to verdict in three steps
Sign up
Continue with Google and your account is ready at once, with 100 free credits.
Submit text
Paste text or upload one file, a PDF with selectable text or a .txt or .md file. Each text costs 1 credit per started 512 tokens, about 400 words.
Read the verdict
For each document you get the likelihood that it was AI-generated, how confident the verdict is, and any reliability warnings.
Coverage
Tested across the AI market
Detection rate on answers from the 12 developers with the most answers in our test, out of 26.
- OpenAI99.6 %
- Google99.8 %
- Anthropic99.8 %
- Meta> 99.9 %
- Mistral99.8 %
- Alibaba99.9 %
- DeepSeek99.9 %
- Reka99.9 %
- Microsoft99.8 %
- 01.AI99.9 %
- Cohere99.9 %
- xAI99.8 %
Late-2025 flagship models
- Anthropic99.6 % of 284
- Google100 % of 633
- xAI99.0 % of 102
- OpenAI98.8 % of 336
Answers were collected up to October 2025. Newer models are not yet benchmarked.
Head to head with Pangram
Within 5 points of Pangram 4 on 17 of 21 results
Pangram is one of the leading AI text detectors. We scored ParaTrace v10-ft-g on the same public tests Pangram reports on, from student essays to novels, news and reviews. It catches 100 % of fully AI-written essays, the same as Pangram, misses fewer AI texts than both Pangram versions on writing by well-known authors, and is within 5 points of Pangram 4 on 17 of the 21 results.
See all 21 results against PangramWithin 5 points
- Against Pangram 417 of 21 results
- Against Pangram 318 of 21 results
Results where ParaTrace is level, ahead, or no more than 5 points behind.
Scientific papers
No false alarms on research papers written before ChatGPT
NeurIPS checked its 2026 position papers for AI-written text with Pangram, after testing it on conference papers from 2022 and 2025. We ran the same check with ParaTrace v10-ft-g. It flags none of the 106 papers from before ChatGPT, exactly like Pangram, and flags 25 of 123 papers from 2025 as at least half AI-written, 20.3 % against 1.0 % for Pangram.
See the field test on scientific papersPapers from before ChatGPT
0of 106 papers flagged by ParaTrace, at any AI score
The same result as Pangram, which flagged 0 of 159.
Good to know
Know the limits
This is a statistical estimate, not proof. Don't use it as the sole basis for decisions about a person.
- English only. Texts in other languages are not supported.
- At strict, 1.0 % of held-out human texts (about 1 in 100) were wrongly flagged as AI in our tests. At lenient, about 5 %. Other kinds of writing can differ.
- Persuasive essays written by adults are a known weak spot. About 1 in 5 (20 %) is wrongly flagged at strict.
- Texts under 50 words give less reliable verdicts. In our tests 4.1 % of short human texts were wrongly flagged.
- Long texts are scored in overlapping chunks and labelled part by part, so a text can come out partly AI-generated. Where the AI part starts and ends is approximate.
Is the AI checker free?
Yes. You can check a few texts a day right on this page, with no account. A free account starts with 100 credits and takes longer texts and files. Each text costs 1 credit per started 512 tokens, about 400 words.
How accurate is ParaTrace?
In our latest benchmark it caught 99.8 % of 25,601 AI-written answers from 90 models, and wrongly flagged 1.0 % of 22,681 held-out human texts. The benchmark report has every result.
Which AI models does it detect?
We tested answers from 90 models by 26 developers, among them the late-2025 flagship models of Anthropic, Google, OpenAI and xAI. Answers were collected up to October 2025, and newer models are not yet benchmarked.
Does a verdict prove that a text was written by AI?
No. This is a statistical estimate, not proof. Don't use it as the sole basis for decisions about a person.
Which languages and files can I check?
English texts only. You can paste a text, or sign in and upload one file, a PDF with selectable text or a .txt or .md file.
What happens to the texts I check?
Free checks without an account aren't stored. Texts you check signed in are stored in your history until you delete them. Don't submit confidential or personal data. The privacy policy has the details.
Can I use ParaTrace from my own software?
Yes. The ParaTrace API checks texts from your own software, and a free key comes with 1,000 credits. To run the detector in your own cloud, see the Enterprise edition.
Check your first text now
Free, in seconds, with no account needed.