Firecrawl Universal Scrape: One /scrape Call for Web Pages, Files and 200+ Data Providers

Firecrawl launched Universal Scrape on October 8, 2026. The idea is simple: your agent calls one endpoint, /scrape, and gets context from web pages, files and a growing set of third-party data providers. This post covers what Firecrawl announced, how the new call works, and what the launch does not yet answer.

What Firecrawl announced

Eric Ciarla, a Firecrawl co-founder, wrote that getting an agent the right context can still mean "building a scraper, parsing a file and connecting another API." Universal Scrape moves more of that work into /scrape. Per the launch post, it covers four kinds of data:

• Web pages and files. Dynamic sites, PDFs and images, plus Word, spreadsheet and presentation files available by URL. Firecrawl handles rendering, retries and parsing.
• People and company profiles. Through providers including Apollo, FullEnrich and Data Legion.
• Financial data. Company financials and market information through providers such as Fiscal.ai and Benzinga.
• Podcasts. Episode and conversation search through Particle, with tools for search, lookup and transcripts.

Alexandria has more than 200 providers. Firecrawl says Universal Scrape "already works with a growing selection" of them and that it adds providers every day. The banner on Firecrawl's site also points to Alexandria and a $75M Series B announcement. We have not verified the funding details beyond that banner.

How the call works

For a web page or supported document, you send the URL as before. For provider data, you pass an alexandria object. Firecrawl's example uses the latest firecrawl SDK and the FIRECRAWL_API_KEY environment variable:

import { Firecrawl } from "firecrawl";

const firecrawl = new Firecrawl({
  apiKey: process.env.FIRECRAWL_API_KEY,
});

const page = await firecrawl.scrape("https://www.firecrawl.dev", {
  formats: ["markdown"],
});

const podcasts = await firecrawl.scrape({
  alexandria: {
    provider: "particle",
    capability: "podcasts/episodes/search",
    options: {
      semantic_search: "AI agents",
      limit: 2,
    },
  },
});

The second call runs Particle's podcast search and returns matching transcript segments. Three details from the post matter in practice:

1. One key, one bill. Provider calls use your Firecrawl key and are billed in Firecrawl credits at each tool's listed price.
2. Terms come first. Before the podcast call works, an org admin must accept Particle's provider terms.
3. Some URLs route automatically. Firecrawl says some supported profile URLs go straight to your configured enrichment providers.

Where you can use it

Firecrawl lists several routes: a plugin for ChatGPT and Codex, a Firecrawl connector in Claude, a Claude Code plugin, the API, MCP and a CLI for terminal work. After you connect your account, you ask the agent to read a URL or find data in Alexandria. For background on the tool itself, see explainx.ai's Firecrawl guide.

What is not answered yet

The launch post is a product announcement. Several questions are open.

• Freshness. One X reader asked whether the response says when each page was fetched, or whether cached data can look identical to fresh data, and said that decides whether they would trust it for financial work. Firecrawl's post does not address this. Treat provider data as unverified for time-sensitive use until you have checked the response fields.
• Pricing. "Each tool's listed price" means costs vary by provider and capability. We have no published price table to compare, so estimate costs with a small test before wiring it into a loop.
• Provider terms and data rights. People and company enrichment data carries its own legal and privacy rules. Accepting a provider's terms is the buyer's responsibility, not something Universal Scrape removes.
• Coverage. "Growing selection" of 200+ providers is not a list of what works today. Check the Alexandria directory for the specific capability you need.

Why it matters

Agent builders have spent 2026 stitching together search, scraping, parsing and enrichment APIs. Exa, Parallel and others compete on search quality, as in explainx.ai's coverage of Exa Snapshot and the Parallel web search eval. Firecrawl's bet is different: be the single entry point, and let the long tail of data sit behind the same call. That reduces integration work, at the cost of depending on one vendor's routing, billing and uptime.

Firecrawl also recently relaunched free keyless search and scrape and open-sourced its pdf-inspector, so the product is widening on both the access and parsing sides.

What to do now

1. Keep your existing /scrape calls. Nothing changes for URL scraping.
2. Pick one provider capability you currently pay for separately and test it through Alexandria on a small sample.
3. Log fetch times yourself if the data feeds decisions, until Firecrawl documents cache behavior.
4. Check robots.txt, site terms and provider terms before scraping or enriching people data.

Sources

• Firecrawl, "Introducing Universal Scrape," Eric Ciarla, October 8, 2026
• Firecrawl announcement post on X, October 8, 2026, including reader replies

Accurate as of October 9, 2026, based on Firecrawl's launch post. Provider list, pricing and cache behavior can change; verify against Firecrawl's current docs before building on them.