anydoc vs MinerU
Rust library converting Word, PowerPoint, Excel, OpenDocument, RTF, EPUB, CSV and PDF to clean Markdown in single-digit milliseconds — Node/Python/WASM bindings, CLI, and an agent skill. — versus — Heavyweight document-to-markdown/JSON parser — PDFs plus Office (docx/pptx/xlsx) through layout analysis and OCR into LLM-ready output for RAG and agentic pipelines. 73k stars, self-hostable.
Both turn office documents into LLM-ready Markdown: MinerU brings ML layout analysis and OCR at pipeline weight; anydoc is a structural converter that answers in milliseconds but skips scans.
| anydoc | MinerU | |
|---|---|---|
| Stars | 11k | 77k |
| Forks | 510 | 6.5k |
| Language | Rust | Python |
| License | MIT | NOASSERTION |
| Last activity | 2 days ago | 3 days ago |
| Topics | ocr | ocr, rag |
| Curated connections | 2 | 12 |
anydoc — the curator's take
The right scope discipline: structural conversion only, brutally fast, one consistent Markdown out regardless of input — and the WASM demo means files never leave the machine. Ships as an agent skill, so agents read office docs without a service. When NOT: scanned pages — there is no OCR in it (that's Firecrawl's hosted Parse upsell), so image-heavy PDFs come back empty; for layout-model parsing of messy documents you still want the heavyweight pipelines.
MinerU — the curator's take
The incumbent when document variety is the problem: beyond PDFs it handles Office formats, with mature layout analysis (reading order, tables, formulas) and a huge user base shaking out edge cases. NOT the lightest option — it's a full pipeline with model downloads and real hardware appetite; for a handful of clean PDFs a smaller tool is faster to stand up. Check the license (NOASSERTION on GitHub — AGPL-family, matters for commercial use).