Integrations / Search

Firecrawl

Turn web pages into clean markdown with Firecrawl, search the web with page content included, map a site's URLs, crawl whole sites and extract structured data with a prompt.

9 tools

Tools

firecrawl.scrape

Read one page as clean markdown, rendering JavaScript first. Can also return a short summary, the page's links, its HTML or a screenshot URL. Cached copies up to two days old are used unless max_age says otherwise.

url*formatsonly_main_contentinclude_tagsexclude_tagswait_formax_agemobiletimeout

firecrawl.search

Search the web, news or images. Returns title, URL and description for each result; with scrape_formats, each web and news result also carries its page as markdown, which costs a scrape's credits per result.

query*limitsourcestbslocationcountryinclude_domainsexclude_domainsscrape_formatstimeout

firecrawl.map

List a site's URLs from its sitemap and links, quickly and without reading the pages. With search, the URLs most relevant to it come first.

url*searchlimitsitemapinclude_subdomainstimeout

firecrawl.start_crawl

Start crawling a site from a URL, reading every page it reaches as markdown. Returns the crawl's ID, a UUID; follow it with get_crawl until its status is completed.

url*promptlimitinclude_pathsexclude_pathsmax_discovery_depthcrawl_entire_domainallow_subdomainssitemapformatsonly_main_content

firecrawl.get_crawl

Get a crawl's status (scraping, completed or failed), its progress and the pages read so far. When next is set, more pages remain: read them with api_get on next's path.

id*

firecrawl.extract

Start extracting structured data from one or more pages with a prompt, a JSON schema or both. URLs may end in /* to cover a whole site. Returns the job's ID, a UUID; follow it with get_extract until its status is completed.

urls*promptschemaenable_web_searchshow_sources

firecrawl.get_extract

Get an extraction's status (processing, completed, failed or cancelled) and, once completed, the extracted data.

id*

firecrawl.api_get

Send a GET request to any Firecrawl API path under https://api.firecrawl.dev/v2 and return the response. For endpoints the named tools don't cover.

path*params

firecrawl.api_request writes

Send a POST, PUT, PATCH or DELETE request to any Firecrawl API path under https://api.firecrawl.dev/v2.

method*path*paramsbody

Tick Allow raw writes on the connection to add api_request, which can send POST, PUT, PATCH and DELETE requests to any Firecrawl API path.

Connect

In the Firecrawl dashboard (firecrawl.dev/app), open API Keys and copy a key, or create one.
Paste it (fc-…) as the connection's API key. Scrapes, searches, maps, crawls and extractions are billed in credits to that team; verifying reads the remaining credits and costs nothing.

Credentials

API key

Firecrawl API docs

Turn one person's AI process into the company's.

Murmurator is invite-only. Already invited? Sign in and build something real.