Firecrawl announced a $75 million Series B led by Smash Capital on 22 September, following a $14.5 million Series A in August 2025, Wowtale reported.
Key points
- Altos Ventures, Nexus Venture Partners, Y Combinator, Freestyle and Offline Ventures also joined the Series B.
- Firecrawl says its Alexandria product combines paid data providers, its own indexes and the live web through one interface.
- In Firecrawl’s internal evaluations, agents using Alexandria scored 21 percent higher on answer quality than agents using built-in web tools.
- The company plans to let more contributors supply information and receive payment based on demand from AI agents.
Smash Capital backs Alexandria’s expansion
Firecrawl said the round also included Altos Ventures, Nexus Venture Partners, Y Combinator, Freestyle and Offline Ventures. It intends to spend the capital on broader Alexandria coverage and payments to researchers, publishers and data providers. Its stated priorities are larger indexes and retrieval systems, direct connections to more original sources, and infrastructure for paying contributors.
The company was founded in 2023 by Caleb Peffer, Eric Ciarla and Nicolas Silberstein Camara, who had previously built the document-chat tool Mendable, Wowtale reported. Firecrawl’s original service collects website content and returns text in markdown or JSON for developers to use through an API. Alexandria extends that service to information obtained directly from providers and from indexes assembled by Firecrawl.
Firecrawl says more than 1.5 million users build with its search, scraping and retrieval tools. Wowtale reported that Zapier integrated the API to let a chatbot read customer websites and generate FAQ answers, while Replit uses it to gather developer documentation. Firecrawl also says Shopify and several large hedge funds use its tools for tasks including lead generation, price monitoring and market research.
Alexandria connects Wikimedia Enterprise and Firecrawl indexes
Firecrawl says Alexandria lets an agent discover and retrieve information from official providers, specialist indexes and the live web through the Firecrawl API. Its agreements with official providers are paid: the company says it already pays Wikimedia Enterprise for direct access to its data, with millions of Wikipedia-data requests passing through the service each month.
The company’s Research Index covers tens of millions of scientific paper abstracts. Its Developer Index holds tens of millions of sources drawn from documentation, READMEs, issues and merged pull requests. A Government Index covers laws, regulations and ordinances. Alexandria also offers connectors for particular websites and uses Firecrawl’s search and scraping tools to reach the live web.
Firecrawl describes Alexandria’s provider network as 88 official data providers, registries and publishers offering 504 capabilities across 28 categories. It puts the number of indexed sources at more than 113 million. The company says Alexandria is available through its MCP, CLI and API.
Firecrawl reports a 21 percent test advantage
In internal evaluations across different fields, Firecrawl says agents using Alexandria scored 21 percent higher on answer quality than agents relying on built-in web tools. The result is a company-run comparison of answer quality, rather than a measure of customers’ spending on the product.
Peffer, Firecrawl’s co-founder and CEO, said the company wants individuals and organisations to earn money when agents use their information. Firecrawl plans a self-service system through which contributors can provide knowledge and be paid according to agent demand. That planned system would extend its existing individual agreements with official providers to a wider group of contributors.
Exa and TollBit pursue adjacent markets
Firecrawl faces companies focused on different parts of the agent-data business. Exa, which sells search tools for AI, raised a $250 million Series C at a $2.2 billion valuation. Its search offering competes with Firecrawl’s tools for finding web content, while Alexandria adds licensed information and Firecrawl’s indexes to the material agents can retrieve.
Publishers have other routes to payment for AI use of their material. TollBit lets publishers retain 100 percent of revenue while charging AI companies per scrape. ProRata.ai shares subscription and advertising revenue from its licensed-content answer service with publishers according to how often their material appears.
