Standard parsing is free during launch. Pay per page only when you need premium.
Run locally. Full parsing capabilities. No limits.
The hosted API with your own key. No infrastructure, no card required.
Everything in Standard, plus OCR on every page — so the parts naive parsers drop come back: tables, images and equations.
Complete RAG pipeline. Graph + hierarchical embeddings.
Interested in dedicated servers, custom configs, or on-premise deployment? Let's talk and work out a solution together.
One page of the source PDF. A 10-page document costs 10× the per-page rate.
Two. The open source parser is free forever, with no limits. And during launch, the hosted API's standard tier is free too — sign up, create a key, no card required.
An OCR pass on every page, grafted onto the standard parse. That is what brings back tables as tables, pictures as image nodes with their bytes, and display math as LaTeX — the same bgraph, with more of the document in it. $0.01 per page at the launch price; $0.02 after.
For teams processing thousands of pages, we offer dedicated infrastructure with flat monthly pricing. Get in touch.
The parser is fully open source — run it locally with no limits. For managed dedicated servers, contact us.