MiSource available
MinerU
Turn complex papers and reports into structured content, including text, tables, and formulas.
Agentic Document AI
DataMosaic
Build document-derived databases with DBMosaic or investigate long visual documents with DocMosaic.
DOCUMENTATION-BASED SELECTION
Why it is a Featured Tool
DBMosaic combines target structures with extraction, verification, and repair; DocMosaic adds cross-page evidence investigation for long visual documents.
Before you choose: Real processing requires deployment and configured models. The project site includes a static demo, and modules maintain separate workspaces.
See the shortlist and criteria →Explore the DataMosaic project site ↗
At a glance
- Input
- PDF for DocMosaic; document collections for DBMosaic
- Output
- Relational records, databases, answers
- Provenance
- Navigable answer evidence; field coordinates need verification
- Scanned pages
- Depends on the parsing backend
- Deployment
- Self-hosted
- Availability
- Source available
- Costs to plan for
- Compute or model API costs
Limitations to check
The project website includes a static demonstration. Real processing requires deployment. Modules do not automatically share files; answer citations do not guarantee field coordinates.
Build a meaningful sample
Include your hardest pages: low-quality scans, missing fields, dense tables, and evidence split across pages. Validate output correctness separately from citation accuracy.