Firecrawl, founded by Caleb Peffer (@CalebPeffer), Eric Ciarla (@ericciarla), and Nicolas Silberstein Camara (@nickscamara_), released a rebuilt Model Context Protocol server on July 30th that it says consumes 50% less context on search, scrape, and interact calls.
https://x.com/firecrawl/status/2082864634764075135?s=46 The release announced on X also adds browser-based OAuth onboarding for people and a keyless route for agents. Firecrawl is packaging those changes as a faster way to give any MCP-compatible AI client access to current web pages without forcing users to install a local server or manually copy credentials into a configuration file.
Firecrawl's founders arrived at the problem while building Mendable, a chat product that ingested documentation. They repeatedly had to solve the same underlying problem: collecting web pages, handling JavaScript and rate limits, cleaning the results, and converting them into data that a language model could use. They eventually separated that infrastructure from the chatbot and built Firecrawl around the API layer.
Peffer previously described that shift in Firecrawl's August 2025 Series A announcement. Firecrawl's about page says the founders also built SideGuide and now operate from San Francisco. The three founders participated in Y Combinator's Summer 2022 batch.
Cutting the MCP context bill
The central claim in Thursday's update concerns the amount of a model's context window consumed each time Firecrawl runs /search
, /scrape
, or /interact
. MCP servers can occupy context with tool definitions and the information returned from each call, leaving less room for the model's instructions, conversation history, retrieved documents, and reasoning.
Firecrawl says the rebuilt MCP cuts that consumption in half on each of the three calls. That is a company benchmark, and the announcement does not establish a universal token reduction across every model and client configuration. The practical goal is clear: let agents perform more web operations before context limits force them to compress, discard, or reload information.
The Firecrawl MCP documentation exposes tools for web search, individual-page scraping, site mapping, crawling, document parsing, autonomous research, and browser interaction. The interact tool can click, navigate, fill forms, and operate dynamic pages after a scrape, extending Firecrawl beyond the read-only retrieval handled by a conventional search API.
Authentication now has three paths. Interactive MCP clients can use OAuth, which sends a person through a browser sign-in rather than asking them to retrieve and paste an API key. Servers, scripts, and other unattended workloads can continue sending a Firecrawl key through an authorization header. A rate-limited keyless endpoint lets agents start with no account or credential.
Firecrawl's current MCP material says keyless access covers Search, Scrape, and Parse, while OAuth or an API key opens the broader tool surface. That boundary gives Firecrawl a low-friction entry point without offering unlimited anonymous access to heavier operations such as crawling and autonomous research.
Firecrawl removes another onboarding step
The keyless component builds on a release Ciarla published on June 16th. Firecrawl initially pitched keyless access as a way for agents, prototypes, workshops, and small projects to use web data before a developer creates an account. The new MCP release brings that distribution strategy together with OAuth and lower context use.
The timing reflects where web-data providers are competing. Firecrawl lists Tavily, Exa, Apify, Bright Data, Browser Use, Browserbase, Jina AI, and others among its alternatives. Search quality and scraping reliability remain core purchasing criteria, but agent developers also have to consider token consumption, authentication, tool schemas, and whether an integration works across multiple AI clients.
Firecrawl is trying to make those operational details part of its advantage. The open-source Firecrawl repository provides the underlying web extraction software, while Firecrawl sells hosted access and higher-volume plans. Firecrawl says it has served more than 5 billion requests and signed up more than 1.25 million developers; those figures are self-reported.
Firecrawl has raised $16.2 million, according to its website. That includes a $14.5 million Series A announced on August 19th, 2025, led by Nexus Venture Partners, with backing from Y Combinator, Zapier, Shopify CEO Tobias Lutke, Postman CEO Abhinav Asthana, and Mux founder Matt McClure.
Lower context consumption gives Firecrawl a direct cost and usability pitch to developers building agents that make repeated web calls. OAuth and keyless access address the distribution side of the same bet: Firecrawl wants its MCP endpoint to become the default route an agent takes when it needs information from the live web.