Drop a PDF. Get structured Markdown - with footnotes, LaTeX, and figure links.
PaperFlow turns raw PDF->Markdown into structured knowledge. Citations become clickable footnotes. LaTeX renders perfectly. Figure references become internal links. Choose a free local parser or connect a cloud API for best quality.
Quick decision guide
Quick selection guide
Standard digital PDF?
+-- Yes -> PyMuPDF Local
`-- No
+-- Need local AI for scans / formulas / layout recovery?
| `-- PaddleOCR-VL-0.9B
+-- Cloud is acceptable and you want the easiest premium path?
| `-- Marker API (Datalab.to)
`-- Need private infrastructure and enterprise control?
`-- Enterprise Marker Self-Hosted
1. Choose parser
Choose by document type, privacy requirements, and setup effort. Every option still gets the same PaperFlow post-processing: footnotes, LaTeX cleanup, figure links, and YAML metadata.
Select a parser
Choose a parser to see which workflow fits your documents.
Pick the option that matches your document type, privacy requirements, and setup tolerance.
2. Upload PDF
Your files never leave your machine. All processing happens locally (or via your configured API). Supports single and batch processing.
Required only for Marker API. PaperFlow will use your key for this run.
or click the button to choose one or more files
3. Status
PaperFlow will poll the backend until the single job or batch run finishes.
4. Markdown preview
Shows the first part of the processed markdown once the job completes.
Preview will appear here after processing completes.
Need PaperFlow deployed in your organization?
Private deployment on your servers · Commercial API · Custom document integrations
Contact us ->