anydoc

anydoc is a document converter published by Firecrawl (GitHub: firecrawl). Its core is written in Rust and works as a library, a command-line tool, or WASM in the browser, and the repository includes an Agent Skill. The README says it turns office documents into LLM-ready Markdown "in single-digit milliseconds, with one consistent output no matter which format goes in." Ordinary conversion happens entirely on your machine, and the Rust library itself makes no network calls. For files that need OCR, such as scanned PDFs, you can opt into a hosted OCR service. Install through npm, pip, or cargo, or add it as a Skill with npx skills add.

Who it's for:Developers who often feed documents to AI or are building document pipelines for agents. It is of no use if you don't write code.

Before you start: Ordinary conversion runs locally and needs no account. Hosted OCR is used only if you choose it, and then the whole document is uploaded to Firecrawl Parse; it cannot send just selected pages, so don't use that option for files you don't want to upload. The README says hosted OCR works without signing up and an API key raises the limits; pricing and data retention terms were not covered in the check, so refer to the official site. The license is MIT. The documents you convert still belong to their authors, and the tool's license does not change that.

Based on the project's public documentation. Not tested by AI Inspiration World.