anydoc

anydoc

Office docs to Markdown for LLM input

WHO USES ITPeople feeding office files to an LLM · RAG and agent developers · Privacy-conscious users

Use in the browser

Converts Word, PowerPoint, Excel, PDF and EPUB to Markdown; run via npx or in the browser

★ 21.5k▲ +40 / 7dlast 14 days · stars per dayhollow = no dataAs of 09-16 21:00 UTC

21,535 stars · checked on GitHub 09-15 19:34 UTC GitHub check 30 h late · incl. +2 observed via GH Archive since then · 7-day +40 observed via GH Archive (as of 09-16 21:00 UTC)

RANKS PRODUCTIVITY #19 As of 09-16 21:00 UTC

anydoc browser demo with the document drop area, the converted Markdown output and Copy / Download .md buttons

TESIGN TAKE

It is a library at heart, but the one-line npx CLI and the browser demo make it usable without code; just remember that a scanned PDF leaves for Firecrawl's servers only when you opt in with --ocr hosted.

AT A GLANCE

WHY IT MATTERS

anydoc parses .doc/.docx, .ppt/.pptx, .xls/.xlsx, OpenDocument, RTF, EPUB, CSV and PDF into one document model and renders through one GFM serializer, so tables, footnotes and equations (as LaTeX) behave the same whatever the input. On its 100-document benchmark it converted in a median 4.4 ms per document and, per the README, was the only tool covering all fourteen formats (quality judged by an LLM). The browser demo runs as WebAssembly, so files never leave the machine.

BUILD FROM THIS

Turn supplier quotes and product sheets (xlsx, docx, pdf) into Markdown with one npx line for a searchable archive or LLM input, or use the WASM package to build a browser-only upload converter with no server.

WHO IT'S FOR

People feeding office files to an LLM: docx, pptx, xlsx to clean Markdown
RAG and agent developers: Node, Python, Rust and WASM bindings plus an Agent Skill
Privacy-conscious users: convert in the browser demo without uploading anything

START IN 5 MINUTES

# npx @firecrawl/anydoc report.docx               # Markdown to stdout
# npx @firecrawl/anydoc slides.pptx -o slides.md  # or to a file
# npm install -g @firecrawl/anydoc
# pip install firecrawl-anydoc

CAVEATS

MIT. Text-based PDFs convert locally, but scanned or image-only PDFs fail with NeedsOcr since there is no built-in OCR; opting into --ocr hosted sends the whole document to Firecrawl Parse (no signup needed, a key raises limits).
Images appear only as alt text in the Markdown, with the raw bytes kept on the document model.
The benchmark corpus is not redistributable and is not in the repo.

RECEIPT

SIMILAR TOOLS

Up to five from PRODUCTIVITY by ★ total: edited entries first, then repository cards; this entry is highlighted. Drawn from the same stored snapshot as the rankings — a comparison, not a recommendation. As of 09-16 21:00 UTC

NAME	★ TOTAL	▲ 7d	LICENSE	PLATFORM	LAST PUSH
Hermes AgentSelf-improving personal agent	★ 246k	▲ +375	MITpermissive	linux · macos · windows · android · cli	2026-09-15
anydocOffice docs to Markdown for LLM inputTHIS ENTRY	★ 21.5k	▲ +40	MITpermissive	cli · web	2026-08-28
immich-app/immichCARD	★ 114k	▲ +82	AGPL-3.0copyleft	—	2026-09-13
stirling-tools/stirling-pdfCARD	★ 91.9k	▲ +54	Licence unconfirmedunconfirmed	—	2026-09-13
sherlock-project/sherlockCARD	★ 91.5k	▲ +64	MITpermissive	—	2026-09-13