Jump to content

Draft:Firecrawl

From Wikipedia, the free encyclopedia
  • Comment: In preparing this draft, I disclose that AI assistance was used as follows: I used an AI assistant to help me draft the text. I checked every citation against its source and reviewed the final wording myself. Juampihz (talk) 13:06, 12 November 2026 (UTC)

Firecrawl
TypePrivate
IndustrySoftware
Founded2022; 4 years ago (2022)
Founders
  • Caleb Peffer
  • Eric Ciarla
  • Nicolas Silberstein Camara
Headquarters
San Francisco, California
,
United States
Key people
  • Caleb Peffer (CEO)
  • Nicolas Silberstein Camara (CTO)
Products
  • Firecrawl
  • Alexandria
Websitewww.firecrawl.dev

Firecrawl is an American software company based in San Francisco that develops a web-data platform of the same name. The platform searches the web, retrieves and crawls websites, and converts their content into formats such as Markdown and JSON for use by large language models and AI agents.[1][2][3] Its core software is published as open-source software, alongside a commercially hosted version.[1] The company is legally SideGuide Technologies, Inc., a Delaware corporation doing business as Firecrawl, and took part in Y Combinator's Summer 2022 batch.[4][5] Its customers have included Shopify, Replit, Zapier, and Canva.[1][6] In September 2026 the company raised a $75 million Series B round and launched Alexandria, a service through which AI agents can retrieve information from third-party data providers, Firecrawl-maintained collections, and the live web.[7][8][9]

History

[edit]

Origins and Mendable

[edit]

Firecrawl was co-founded in 2022 by Caleb Peffer, Eric Ciarla, and Nicolas Silberstein Camara,[1] who had met as roommates at the University of New Hampshire. After graduating they ran an Instagram community that taught programming, which grew to about 50,000 developers and led to a programming-education product and their acceptance into Y Combinator, which invested $500,000.[6] During the program the team moved from consumer to business software, and after the release of ChatGPT it launched Mendable, a chatbot that answered questions from companies' technical documentation and was used by companies including Snapchat and DoorDash.[6]

Open-source launch

[edit]

To feed Mendable with web data, the team built an internal web scraper. According to Camara, they packaged it over a weekend and released it as open-source software on a Monday, and it spread on Twitter and LinkedIn while the founders were flying to San Francisco. The response led them to shut down Mendable and focus entirely on the scraper, which became Firecrawl.[6] The project launched in April 2024.[10][11] Its GitHub repository passed 34,000 stars by early 2025[2] and approached 50,000 by August 2025;[1] by September 2026 Exame reported more than 175,000 stars, placing it among the 100 most-starred projects on GitHub.[9]

In February 2025 the company, then seven people, drew press attention for a job advertisement seeking to hire an AI agent rather than a human, which Peffer described as "equal parts PR stunt, experiment"; a second round of agent listings in May 2025 carried a $1 million budget, but no agent had been hired by August.[12][13][1]

Series A, search, and version 2

[edit]

In August 2025 Firecrawl announced a $14.5 million Series A round led by Nexus Venture Partners. TechCrunch reported that Shopify chief executive Tobias Lütke joined the round after the founders noticed he had signed up for the product through its self-service portal and emailed him; Shopify later approached the company for an enterprise contract.[1] At the time the company said it had 350,000 developers using the product, was profitable, and counted Shopify, Replit, Zapier, and several large hedge funds among its customers. It had also just added web search to its API.[1] Version 2 of the software was released alongside the funding.[14]

Growth and Alexandria

[edit]

By February 2026, Época Negócios reported that the company had about 20 employees, 12,000 paying customers, and 600,000 users, with clients including Canva, Shopify, OpenAI, and Nvidia.[6] On September 22, 2026, Firecrawl announced a $75 million Series B round led by Smash Capital and launched Alexandria.[7][8][9] The company said at the time that more than 1.5 million developers and 150,000 companies used its products, including Shopify, Apple, Lovable, and Canva.[9]

Funding

[edit]
DateRoundAmount (US$)Lead investorOther participants
2022Accelerator$500,000Y Combinator—
August 2025Series A$14.5 millionNexus Venture PartnersY Combinator, Tobias Lütke
September 2026Series B$75 millionSmash CapitalAltos Ventures, Nexus Venture Partners, Y Combinator, Freestyle, Offline Ventures

Sources: Y Combinator investment;[6] Series A;[1] Series B.[7][8][3] The company said much of the Series B funding would go toward adding third-party data providers to Alexandria.[7]

Products and services

[edit]

Firecrawl

[edit]

Firecrawl is offered as a hosted commercial API and as open-source software that can be self-hosted.[1][2] The main source-code repository is licensed under the GNU Affero General Public License, version 3.[2][10]

Its core functions retrieve the content of a single page ("scrape"), follow links to collect the pages of an entire website without requiring a sitemap ("crawl"), list the URLs on a site ("map"), and run web searches that can also return the content of the result pages ("search").[2][3][15] Output is returned as Markdown, HTML, or structured JSON, and the service renders pages that rely on JavaScript before extracting their content.[2][3] SiliconANGLE described the search function in 2026 as a search engine designed for AI agents that can filter results by attributes such as publication date and file type.[7]

Version 2, released in August 2025, added a caching layer that stores previously retrieved pages as an index, and a JSON mode in which users describe in a natural-language prompt the information they want extracted and its format.[14] According to SiliconANGLE, the platform can also wait for pages that load in stages, perform interactions such as scrolling, clicking, and submitting forms based on user prompts, extract content from PDF and Microsoft Word files, and monitor selected web pages for changes, for example to track updates to a competitor's pricing page.[7] An "agent" endpoint, which replaced an earlier "extract" endpoint, accepts a prompt and searches and navigates websites to gather the requested data when the relevant URLs are not known in advance.[16]

Besides the API, Firecrawl publishes software development kits for languages including Python, Node.js, Go, Java, Ruby, and Rust;[17] a Model Context Protocol (MCP) server;[18] and a command-line interface with accompanying agent "skills", instruction packages that let AI coding assistants call the service.[19][20]

Alexandria

[edit]

Alexandria launched on September 22, 2026, alongside the Series B announcement. It gives AI agents a single interface for finding and retrieving information from third-party data providers, from collections maintained by Firecrawl, and from the live web through Firecrawl's own search and scraping tools, and can be accessed through an API, an MCP server, or the command-line interface.[7][8][3][9] At launch, Firecrawl's collections included an index of tens of millions of scientific paper abstracts, a similarly sized index of software documentation and code, and government documents such as court records and U.S. Securities and Exchange Commission filings.[7][9] Listed third-party sources included company financial data from Fiscal.ai, podcast transcripts from Particle, and Wikimedia Enterprise, whose paid API Firecrawl had begun using in March 2026 in place of scraping Wikipedia pages.[8][21] Searching for a data source is free, while retrieving data costs the credits listed for each source.[8]

Firecrawl said it intended to pay individuals, publishers, and other organizations when AI agents use their content. Search Engine Watch reported that at launch the company paid some providers under individual agreements, while the self-service system for new contributors was still a waitlist and payout rates had not been published.[8] Wowtale compared the service with content-licensing companies such as TollBit and ProRata.ai.[3]

Customers and integrations

[edit]

TechCrunch named Shopify, Replit, and Zapier as customers in 2025.[1] According to Wowtale, Zapier used the API to let its chatbot read a customer's website and generate FAQ answers, and Replit used it to collect and organize developer documentation.[3] Época Negócios reported that Shopify used Firecrawl in onboarding, extracting data from merchants' existing online stores to help migrate them to its platform.[6]

Firecrawl is available as a built-in connector in the AI app builders Lovable and Replit,[22][23] as a plugin for ChatGPT,[24] and as an app on Zapier.[25] InfoWorld noted integrations with the LLM application frameworks LangChain and LlamaIndex.[2]

Reception

[edit]

In August 2025, TechCrunch noted that AI web crawlers had gained "a somewhat dubious reputation", largely because of operators that ignore robots.txt files, and reported that Firecrawl's founders said they were developing tools to let website owners and publishers be paid when AI systems use their content.[1] A September 2025 VentureBeat article on companies choosing between building their own scrapers and buying a service included interviews with customers who had evaluated Firecrawl against in-house tools and competitors such as Exa.[14] InfoWorld published an overview of the open-source project in June 2025,[2] and the Brazilian business publications Época Negócios and Exame covered the company's growth, including its expansion in Brazil.[6][9]

References

[edit]
  1. 1 2 3 4 5 6 7 8 9 10 11 12 Bort, Julie (August 19, 2025). "AI crawler Firecrawl raises $14.5M, is still looking to hire agents as employees". TechCrunch. Retrieved November 12, 2026.
  2. 1 2 3 4 5 6 7 8 MSV, Janakiram (June 19, 2025). "Firecrawl: Easy web data extraction for AI applications". InfoWorld. Retrieved November 12, 2026.
  3. 1 2 3 4 5 6 7 Kim, Taehyun (September 23, 2026). "Firecrawl Raises $75M Series B, Launches Alexandria". Wowtale. Retrieved November 12, 2026.
  4. ↑ "Terms of Service". Firecrawl. October 7, 2026. Retrieved November 12, 2026.
  5. ↑ "Firecrawl". Y Combinator. Retrieved November 12, 2026.
  6. 1 2 3 4 5 6 7 8 Julio, Rennan (February 3, 2026). "A startup fundada por um brasileiro nos Estados Unidos que encantou Canvas, Shopify e até OpenAI e Nvidia" [The startup founded by a Brazilian in the United States that won over Canva, Shopify, and even OpenAI and Nvidia]. Época Negócios (in Portuguese). Retrieved November 12, 2026.
  7. 1 2 3 4 5 6 7 8 Deutscher, Maria (September 22, 2026). "Web scraping startup Firecrawl closes $75M investment". SiliconANGLE. Retrieved November 12, 2026.
  8. 1 2 3 4 5 6 7 Ashmere, Martin (September 22, 2026). "Firecrawl wants to pay publishers when AI agents use their content". Search Engine Watch. Retrieved November 12, 2026.
  9. 1 2 3 4 5 6 7 Gonçalves, Guilherme (September 22, 2026). "Startup de IA capta R$ 400 milhões para criar biblioteca global de conhecimento para agentes" [AI startup raises R$400 million to build a global knowledge library for agents]. Exame (in Portuguese). Retrieved November 12, 2026.
  10. 1 2 "firecrawl/firecrawl". GitHub. Retrieved November 12, 2026.
  11. ↑ Ciarla, Eric. "Firecrawl: Open source crawling and scraping for AI-ready web data". Y Combinator. Retrieved November 12, 2026.
  12. ↑ Bort, Julie (February 14, 2025). "A job ad for Y Combinator startup Firecrawl seeks to hire an AI agent for $15K a year". TechCrunch. Retrieved November 12, 2026.
  13. ↑ Bort, Julie (May 17, 2025). "Y Combinator startup Firecrawl is ready to pay $1M to hire three AI agents as employees". TechCrunch. Retrieved November 12, 2026.
  14. 1 2 3 Kerner, Sean Michael (September 8, 2025). "Cut AI data prep time by 33%: Why enterprise teams are ditching DIY web scrapers". VentureBeat. Retrieved November 12, 2026.
  15. ↑ "Map". Firecrawl Docs. Retrieved November 12, 2026.
  16. ↑ "Agent". Firecrawl Docs. Retrieved November 12, 2026.
  17. ↑ "SDKs". Firecrawl Docs. Retrieved November 12, 2026.
  18. ↑ "firecrawl/firecrawl-mcp-server". GitHub. Retrieved November 12, 2026.
  19. ↑ "Introducing Firecrawl Skill and CLI: The Complete Web Data Toolkit for Agents". Firecrawl. Retrieved November 12, 2026.
  20. ↑ "Introduction". Firecrawl Docs. Retrieved November 12, 2026.
  21. ↑ "Firecrawl Drops Wikipedia Scraping for Wikimedia Enterprise APIs". Wikimedia Enterprise. March 17, 2026. Retrieved November 12, 2026.
  22. ↑ "Firecrawl". Lovable Documentation. Lovable. Retrieved November 12, 2026.
  23. ↑ "Integrations overview". Replit Docs. Replit. Retrieved November 12, 2026.
  24. ↑ "Firecrawl plugin for ChatGPT". OpenAI. Retrieved November 12, 2026.
  25. ↑ "Firecrawl Integrations". Zapier. Retrieved November 12, 2026.
[edit]

Category:Software companies of the United States Category:Software companies established in 2022 Category:Companies based in San Francisco Category:Y Combinator companies Category:Artificial intelligence industry in the United States Category:Web scraping Category:Free software companies