Webclaw

webclaw is a web scraping API that converts public web pages into Markdown, JSON and structured data for LLMs and AI agents. It is offered as a hosted REST API, a command line tool, an MCP server with twelve tools for clients such as Claude, Cursor, Windsurf and Codex, and an open source core that can be self hosted. The core is a Rust project under AGPL 3.0. The API covers scrape, crawl, map, batch, extract, summarize and diff endpoints, with nine output formats including Markdown, plain text, JSON, links and raw HTML. It attempts JavaScript rendering and handling of protected pages, supports prompt to schema extraction, page level question answering, YouTube transcript extraction and per request cache settings. Compatible routes for the Firecrawl v2 scrape, crawl and search interfaces are provided to reduce migration work. Typical uses are feeding RAG pipelines, giving agents live web access, monitoring prices and running deep research jobs. Pricing is credit based, with one credit per page for every endpoint. The free plan gives 500 credits a month with no card. Starter is $19 per month for 5,000 credits, Growth is $49 for 25,000, Pro is $99 for 100,000 and Scale is $399 for 500,000, with yearly billing offering two months free. Credit packs start at $20 and purchased credits roll over. Research jobs are counted separately as monthly runs. A Dedicated tier with single tenant deployment is available by contacting the vendor.