Async by default
POST /v1/documents returns 202 and a job id. OCR can take minutes; we never hold the HTTP request open.
Structured JSON from PDFs, billed per page
A developer API for native and scanned PDFs. Upload a file, poll a job, download JSON. Processed in the EU on a simple VPS — not a hyperscaler lock-in.
Create an account Read the API docsPage credits
Transparent monthly credits. A succeeded parse debits the page count from the document. Soft-fail with 402 when the balance is empty.
Your engine, your data
The parser is our open engine (pdf_parser.extract()). Source files are purged after 7 days by default.