Context.dev combines all scraping interfaces into one: web pages, PDFs, and Office files handled in one go.

Beating AI News Flash: YC S26 startup Context.dev launches Scrape Anything, merging previously fragmented web scraping APIs into a single interface. Developers only need to provide a webpage or online file URL and select the desired format to get Markdown, HTML, screenshots, page images, raw files, or structured fields all at once.

In addition to handling regular webpages, it currently supports about 40 file formats including PDF, DOCX, PPT, and XLSX. JavaScript rendering, proxy rotation, anti-scraping handling, and retries are all managed by Context.dev, so developers don't need to maintain their own browser and proxy infrastructure. These capabilities already existed before; this update mainly unifies different inputs and outputs into the same /v1/web/scrape endpoint.

A standard scrape costs 1 credit, and requesting multiple output formats still shares this single base charge; using browser operations costs 2 credits, and PDF OCR is billed separately based on the number of pages recovered.