Web scraping startup Firecrawl closes $75M investment
SiliconANGLE Maria Deutscher
Firecrawl just raised $75M to help AI agents scrape the web. It’s now betting that one API can replace a mess of custom connectors and brittle scraping code.
Based on reporting by SiliconANGLE, Maria Deutscher — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Firecrawl has picked up $75 million in Series B funding to push its web-scraping platform for AI agents further into the market. Smash Ventures led the round, with Y Combinator, Altos Ventures, Nexus Venture Partners, Freestyle and Offline Ventures also backing it.
The pitch is simple enough: AI agents need web data, but the web is messy. Pages load in stages, some content only appears after a click or scroll, and documents like PDFs and Word files can trip up standard scrapers. Firecrawl says its tools are built around those pain points. Its search engine is tuned for agent use, letting developers filter by things like creation date and file type instead of browsing the way humans do.
Then there’s the part that sounds almost mundane until you’ve tried to wire it together yourself. Firecrawl says its platform can wait for pages to finish loading, trigger clicks, submit forms and pull out data from files that aren’t friendly to conventional scraping. Developers can also steer the software with prompts that describe what it should do. The company also says it can monitor pages for changes, which could be useful for tracking something as plain as a competitor’s pricing page.
Alongside the funding, Firecrawl introduced Alexandria, a cloud service that mixes web data with third-party sources and datasets curated by the company. One dataset has tens of millions of scientific paper abstracts. Another has a similar number of code and documentation files. Firecrawl’s argument is that this gives AI agents a way to fill gaps without forcing developers to build a separate connector for every source.
The new money is going toward expanding Alexandria, especially by adding more third-party data providers. Firecrawl also plans to launch a self-service content licensing system soon, which should make that process less painful. The company says the goal is still the same: make web data easier for agents to use, without making developers stitch everything together by hand.
My take — AI-written commentary, not fact-checked reporting
Firecrawl is selling the rare thing AI buyers actually want: less plumbing. That matters more than another glossy agent demo, because most of the real work in AI is still glue code and cleanup. The smart money in this space is on the boring tools that make the flashy ones usable.
Read more about this at: SiliconANGLE