PageSift is a self-hosted stand-in for the Firecrawl API: one Rust binary that turns web pages, PDFs and Office files into Markdown or JSON for agents.
Yes, PageSift (self-hosted, Firecrawl-compatible web crawler and document parser in one Rust binary) is open source under the Apache-2.0 license.
Free under Apache-2.0; commercial support and licensing are offered (docs/COMMERCIAL.md), price not published. No managed cloud. (free).
Yes, PageSift (self-hosted, Firecrawl-compatible web crawler and document parser in one Rust binary) can be self-hosted (the source is available under the Apache-2.0 license).
**PageSift** (YOLOW92/pagesift) is an **Apache-2.0** web crawler and document parser shipped as a single Rust binary. It converts URLs, PDFs (with OCR), Office documents, email and archives into clean Markdown or JSON.
It exposes a **Firecrawl v1/v2-compatible API**, Jina Reader-style /r/ and /s/ endpoints, an MCP server and an agent endpoint, with SDKs for Python, Node, Go, Java, C#, Ruby and PHP. The crawler reads sitemaps, honors robots.txt and applies per-host limits. Install with cargo install pagesift, pip install pagesift, npm install @pagesift/node or the Docker image ghcr.io/yolow92/pagesift.
Account-research and enrichment pipelines often pay per page for Firecrawl-style scraping. PageSift keeps the same API shape on your own server, so existing calls can point at it. It is four weeks old with one maintainer, so test it beside your current provider before switching.