Scrape URL with Firecrawl

scrapeUrlWithFirecrawl loads one URL through Firecrawl and returns the requested LLM-ready content and page metadata. Inputs accept literal values or compatible signals. An optional second argument { account } selects the Firecrawl binding.

Example

automations/scrape-release-notes.automation.ts

import { automation, onHttpRequest } from "automate.ax"
import { scrapeUrlWithFirecrawl } from "automate.ax/firecrawl"

export default automation("Scrape release notes", () => {
  onHttpRequest({ scope: "automation" })

  scrapeUrlWithFirecrawl({
    url: "https://example.com/releases",
    formats: ["markdown", "links", "summary"],
    onlyMainContent: true,
  })
})

Inputs

Input	Type	Required	Default	Description
`url`	`string`	Yes	—	Complete page or public document URL.
`formats`	`(string \| object)[]`	No	`['markdown']`	Content formats such as `markdown`, `html`, `rawHtml`, `links`, `images`, `summary`, `screenshot`, `json`, `changeTracking`, `attributes`, `branding`, `product`, `menu`, `audio`, or `video`. Configurable format objects accept screenshot, JSON Schema, change-tracking, attribute, question, and highlight options.
`onlyMainContent`	`boolean`	No	`true`	Remove page chrome such as navigation, headers, and footers.
`includeTags` / `excludeTags`	`string[]`	No	None	Include or exclude selected HTML tags, classes, or IDs.
`headers`	`Record<string, string>`	No	None	Headers sent to the target page, such as a cookie or user agent.
`actions`	`object[]`	No	None	Wait, click, write, press, scroll, execute JavaScript, take screenshots, scrape intermediate state, or create PDFs before extraction.
`waitFor`	`number`	No	`0`	Additional milliseconds to wait for dynamic content.
`timeout`	`number`	No	`60000`	Provider scrape timeout from 1,000 through 300,000 milliseconds.
`mobile`	`boolean`	No	`false`	Emulate a mobile device.
`location`	`{ country?: string; languages?: string[] }`	No	US	Proxy location and preferred languages.
`proxy`	`'basic' \| 'stealth' \| 'enhanced' \| 'auto'`	No	`auto`	Proxy tier. Enhanced or automatic retries can consume extra credits.
`maxAge` / `minAge`	`number`	No	Provider defaults	Control use of Firecrawl’s indexed cache in milliseconds.
`storeInCache` / `lockdown`	`boolean`	No	Provider defaults	Allow cache storage, or require cache-only retrieval.
`removeBase64Images` / `blockAds` / `skipTlsVerification` / `fastMode`	`boolean`	No	Provider defaults	Provider-native scrape behavior.
`parsers`	`(string \| FirecrawlPdfParser)[]`	No	PDF parser	File parser configuration, including PDF OCR mode and page limits.
`profile`	`{ name: string; saveChanges?: boolean }`	No	None	Persistent browser profile shared by name.
`redactPII`	`boolean \| FirecrawlRedactionOptions`	No	`false`	Redact selected personal-information categories.
`threatProtection`	`object`	No	Account policy	Enterprise threat-protection overrides.
`auditMetadata`	`{ username: string }`	No	None	User identity attached to Firecrawl audit logs.

Headers can contain credentials. Send them only to URLs you trust.

Output

The action returns Firecrawl’s document shape. Optional results include markdown, html, rawHtml, summary, highlights, json, links, images, screenshot, audio, video, attributes, browser actions, changeTracking, branding, product, and menu. Page details remain under metadata, including sourceURL, url, title, description, language, statusCode, scrapeId, contentType, cacheState, creditsUsed, and error when Firecrawl supplies them. Use a JSON format object with prompt or schema when you need structured extraction. The returned json value is JSON and should be narrowed to your expected shape in automation code.

Get started

Guides

Concepts

Management API

Reference

Scrape URL with Firecrawl

Example

Inputs

Output

​Example

​Inputs

​Output

Example

Inputs

Output