Pick one page. Read it free.
Judge the OCR on your own document before you install anything. Drop a PDF or TIFF with 3 pages or 3,000, pick one page and get its text with place and confidence, as JSON or Markdown.
1 File
2 Pick a page
0 pages in this file Read here, free: 1 page
The API also takes multi-page TIFF; each TIFF directory is one page. Here a TIFF is read page by page.
A scanned double-page spread is one image, so it is one page. Pick it like any other.
Click a preview first.
Calling the API from your own code? Build your API request
3 Result
Page
The test reads with a pipeline (the one you chose). A low confidence is a pointer: look here first. A high one is a guide, not a guarantee.
What happens to your file
- Your file is opened inside this tab. The previews are drawn by your browser, with code served from this site. No other site is contacted.
- No key, no account. One page per request, with a limit per network address.
- The page is deleted when the job ends. It is not logged and not used for training.
- Send only pages you may give to a processor. See the privacy notice.