Firecrawl wants to pay publishers when AI agents use their content

Firecrawl has launched Alexandria, a service that connects AI agents to data providers and searchable collections through one interface. The company announced the launch on September 22 alongside a $75 million Series B led by Smash Capital.

Firecrawl says it plans to pay individuals, publishers and other organizations when AI agents use their knowledge. It already pays some providers through individual agreements, but the self-service system for new contributors has yet to open. Its public provider sign-up page is a waitlist.

AI agents can check the price before retrieving data

Alexandria lets an agent find a relevant data source, inspect what it offers and retrieve information from it. Firecrawl lists company financials from Fiscal.ai and podcast transcripts from Particle among the sources, alongside its own research and developer indexes.

The documentation separates discovery from retrieval. Finding tools is free; running one costs the credits listed for that tool. Each listing describes its provider, required inputs, expected response and price. Some providers also require an organization administrator to accept their terms before access is allowed.

Firecrawl’s Wikimedia deal predates Alexandria

The announcement cites Wikimedia Enterprise as an existing paid provider. Wikimedia announced that partnership on March 17, when it said Firecrawl was handling 2 million to 3 million requests for Wikipedia data each month.

Wikimedia described the previous process as repeatedly scraping pages using headless browsers and parsing their HTML. The planned replacement was direct access through its On-demand API, which supplies data in a consistent format. Firecrawl now says it pays for that direct access.

Wikimedia shows what an AI customer pays for

Wikimedia’s content remains available under free licenses. Its paid Enterprise service offers higher-volume access, more frequent updates and service guarantees. The arrangement shows how a provider can charge for making information easier to retrieve and maintain, even when the underlying material is already freely available.

That gives publishers a concrete service to consider offering: structured, maintained data that saves an AI developer the work of repeatedly extracting it from webpages. Whether it pays for that work depends on demand and the contributor’s share of the fees.

Publishers faced a similar information gap in Google’s AI contribution pilot, which SEW covered earlier this month. Alexandria’s public launch materials do not publish payout rates or a revenue split for new contributors. Developers can inspect what a data request costs, but prospective providers cannot yet use those prices to calculate what they would earn.