AnyDoc
firecrawl/anydocReviewed by Radar
What it does
anydoc converts office documents (Word, PowerPoint, Excel, OpenDocument, RTF, EPUB, CSV and text-based PDF) into Markdown, a simple text format that AI tools and websites read easily. It is made by Firecrawl and is aimed mostly at developers and AI agents, through a command-line tool and libraries for Node.js, Python, Rust and the browser. Anyone can try it on a demo web page that converts files on your own computer. Scanned PDFs need optional OCR, which sends the file to Firecrawl's hosted service.
How easy it is to install
A · download and open · Windows
The demo page at firecrawl.github.io/anydoc converts files in the browser with no install; the command-line tool instead needs Node.js (npx) and the libraries need Node.js, Python or Rust.
Numbers on GitHub
- Stars
- 23,281on Oct 31, 2026
- Downloads
- 7,208288 this week
- Growth
- +150stars in 5 days
Stars over time
Facts
- Runs on
- Web, Windows, macOS, Linux
- You need to install
- Nothing extra
- To build from source
- Rust
- Disk space
- 8 MB–24 MB · from the installer
- Accounts or keys
- Firecrawl (hosted OCR for scanned PDFs)
- Cost
- Free · MIT license. Optional hosted OCR through Firecrawl Parse works without signup; a FIRECRAWL_API_KEY gives higher limits.
- License
- MIT (Permissive)
- Can replace
- FineReader
- Last update
- Oct 23, 2026
Risk signals (2)
- Infooptional upload to vendor
README.md: with OCR turned on, "Only documents that need OCR leave the machine, and the whole document goes, since Parse has no page selection" (sent to Firecrawl Parse).
AI review
- Infovendor run benchmark
README.md: the comparison table was run by the authors with an AI judge, and "the corpus is not redistributable and is not in the repo", so the scores cannot be reproduced by others.
AI review
Evidence is shown in its original language.
What Radar could not check
- How faithful the Markdown is for complex files (big spreadsheets, slides with many images, unusual layouts).
- Whether the demo page handles large files as well as the command-line tool.
- What Firecrawl does with documents sent for OCR, and the limits without an API key.
- What the npx command downloads on first run (it fetches a prebuilt binary for your platform).