PDF AI Detector: Check a PDF or Word File for AI
Updated
This PDF AI detector lets you upload a document instead of copying text out of it. Upload a .pdf or .docx file up to 8 MB (or a .txt or .md file), and the text is extracted and checked for signs of AI writing, with a probability score and highlighted sentences. Here is what works, and what to watch for with different files.
Limitations: Every result is a probability that can be wrong, not proof of how something was written. How results can be wrong →
3 free scans in total, shared across all five tools (not per day or per month). No signup needed to start.
How the PDF AI Detector Handles Files
- PDF: text is extracted from the PDF's text layer on our server.
- Word (.docx): text is extracted from the document on our server; older .doc files aren't supported.
- Plain text and Markdown: .txt and .md files are read directly in your browser.
- Limits: files up to 8 MB; the check then needs at least 300 words and uses up to 150,000 characters.
Scanned PDFs and Images of Text
A scanned PDF is a picture of pages, with no text layer, so there is nothing to extract and the check can't run. If you can select and copy words in your PDF reader, the file has a text layer and will work.
If you only have a scan, you would need to run it through optical character recognition (OCR) in another app first. Check the recognised text carefully, since OCR errors can affect the result.
Clean Up Before You Check
Documents often contain more than the writing you care about: cover pages, tables of contents, headers and footers repeated on every page, reference lists, tables and captions. Extracted together, these can distort the score.
For the clearest result, paste just the body text, or upload a copy of the document with the extras removed. For long reports, check one section at a time so you know where flagged sentences came from.
Privacy and Uploaded Files
Uploaded PDFs and Word files are parsed on our server to get the text, and the text is then sent to the detection provider named on the How It Works page. If you are signed in, scan history keeps the file name and the result. Avoid uploading documents with personal or confidential information you aren't allowed to share.
Frequently asked questions
No. Scanned PDFs have no text layer to extract. Run OCR in another app first, then paste or upload the recognised text.
No. Upload a .docx, .pdf, .txt or .md file. Save older .doc files as .docx first.
8 MB per file. The check itself uses up to 150,000 characters of text.
Headers, footers, tables and references extracted with the body text can distort the score. Check the body text on its own.