---
type: tool
url: https://github.com/firecrawl/anydoc
category:
  - "[[Document Ingestion]]"
summary: "Office/PDF/EPUB/CSV → clean Markdown for agents"
added: 2026-10-16
---

## Features
- Office/PDF/EPUB/CSV → clean GFM Markdown (14 formats, single output model)
- Content-based format detection (magic bytes, not extensions)
- Embedded asset extraction via document model (`to_document`)
- Hard resource limits & fuzzing for untrusted inputs
- Agent Skill + CLI/Node/Python/WASM bindings (`npx skills add firecrawl/anydoc`)
- Optional hosted OCR fallback for scanned PDFs (`--ocr hosted`)

## Notes