Firecrawl Raises $75M for Alexandria AI Knowledge Library

Firecrawl has raised a $75M Series B led by Smash Capital, with Altos Ventures, Nexus Venture Partners, Y Combinator, Freestyle, and Offline Ventures participating. The San Francisco company announced the round on September 22, 2026, alongside Alexandria, a knowledge library that connects AI agents to official data providers, specialized indexes, custom connectors, and the live web.

The financing is about more than crawling additional pages. Firecrawl is trying to build a retrieval layer that can discover a source, explain what it contains, make it accessible through a common interface, and compensate the people or organizations that supply the knowledge.

That matters because stronger models cannot reason over information they never receive. As agents take on research, coding, compliance, and commercial work, source coverage, rights, freshness, and traceability become part of product quality rather than an invisible plumbing problem.

What Firecrawl Announced

The $75M Series B gives Firecrawl a new lead investor in Smash Capital while bringing back Nexus Venture Partners and Y Combinator. Altos Ventures, Freestyle, and Offline Ventures also joined the round. Firecrawl did not disclose a valuation, and the public source record does not establish a complete lifetime funding total.

Firecrawl previously announced a $14.5M Series A in August 2025 led by Nexus Venture Partners. Y Combinator, Zapier, Shopify CEO Tobias Lütke, Postman CEO Abhinav Asthana, and Mux founder Matt McClure participated in that financing. The disclosed Series A and Series B amounts total $89.5M, though any earlier financing remains outside that calculation.

The company was built by co-founders Caleb Peffer, Nicolas “Nick” Silberstein Camara, and Eric Ciarla, who currently serve as CEO, CTO, and CGO. Their earlier product, Mendable, exposed the recurring difficulty of extracting clean, reliable information from documentation and the broader web. Firecrawl turned that internal problem into search, scraping, crawling, parsing, mapping, and browser-interaction tools for developers.

Alexandria Expands the Retrieval Layer

Alexandria combines the live web with official providers and Firecrawl's Research, Developer, and Government indexes. Instead of asking developers to maintain separate integrations for each financial database, scientific corpus, government repository, publisher archive, or website, Firecrawl wants agents to discover and retrieve those sources through a consistent layer.

Firecrawl says Alexandria covers more than 113M indexed sources and is available through its MCP, CLI, and API. The product can connect to scientific-paper abstracts, code documentation, READMEs, issues, merged pull requests, laws, regulations, and other primary material. Those figures and coverage descriptions are company-reported, and the exact provider and capability counts were changing on Firecrawl's website on launch day.

The company also reported a 21% improvement in answer quality over built-in web tools across an internal evaluation of 845 tasks. Firecrawl says it used the same model and prompts and relied on blind AI judging. The result is not an independent benchmark, but it identifies the variable Firecrawl wants buyers to notice: an agent's answer can improve when the retrieval system reaches a broader and more appropriate source set.

Why Source Economics Matter

The open web is only one portion of the information an agent may need. Financial data, publisher archives, scientific research, professional expertise, regulatory material, and proprietary datasets often come with access rules, pricing, update schedules, and formats that do not line up neatly. Leaving those sources out can make an answer incomplete even when the model performs exactly as designed.

Firecrawl already pays official providers, including Wikimedia Enterprise, through direct agreements. Alexandria extends that idea by giving providers a route into agent workflows and, eventually, a way to earn when their knowledge is used. Firecrawl says it plans to open a self-service system for individuals, creators, researchers, publishers, and organizations that want to contribute knowledge.

That creates a harder business than web scraping alone. Firecrawl must make incompatible sources usable without erasing their provenance, build pricing that works for both developers and suppliers, and help agents distinguish access from authority. A source appearing in a library does not make every claim inside it correct, current, or appropriate for every task.

What the Series B Will Fund

Firecrawl says the capital will improve search, deepen its indexes and retrieval systems, connect more first-party sources, and expand the payment infrastructure behind Alexandria. That work spans product engineering, data partnerships, source governance, and the commercial systems required to meter and compensate usage.

The company's adoption gives it a meaningful starting point. Firecrawl says more than 1.5M users build with its tools, while its current website lists more than 150,000 companies and displays customers including Shopify, Canva, Apple, DoorDash, Sierra, and others. These are company-reported adoption signals rather than audited operating results, but they explain why data providers may care about joining the network.

Smash Capital is investing as Firecrawl moves from a developer utility toward infrastructure for a knowledge market. The round gives Firecrawl enough capital to negotiate with larger providers, support heavier retrieval workloads, and build contributor economics before a clear standard for agent-mediated knowledge access has settled.

What This Signals for AI Infrastructure

AI infrastructure spending has concentrated heavily on models, chips, and compute. Firecrawl's financing points toward another constraint: the systems that decide which evidence reaches an agent, under what terms, and with what record of origin. Retrieval is no longer just a preprocessing step when the agent is expected to act on the result.

For developers, Alexandria promises fewer bespoke integrations and broader coverage. For data providers, it offers a possible distribution and compensation channel. For customers, the important questions will be less theatrical: whether sources stay current, whether rights are clear, whether costs remain predictable, and whether an answer can be traced back to material worth trusting.

Firecrawl has raised the money to pursue those answers at scale. The Series B will be judged by whether Alexandria can turn a large collection of sources into a reliable exchange among agents, developers, and the people whose knowledge makes the system useful.