In short. Firecrawl Search combines web search with optional page retrieval, using scrapeOptions to return formats such as Markdown and links.

Definition: As of 17 October 2026, Firecrawl Search is an application programming interface (API) that returns web search results and optionally retrieves the content of the resulting pages in the same request.

Key facts

Firecrawl's Search documentation describes adding scrapeOptions to retrieve page content alongside search results.
The Search API reference groups results by source type, including web, news and images; web search is the default.
Firecrawl's Scrape documentation distinguishes the API request status from the HTTP status returned by the target website.

How does search combine with page retrieval?

Firecrawl's Search guide documents sending a query to /search and adding scrapeOptions for page retrieval. Scraping means fetching a page and extracting its content. The following request matches the guide's example and requests page text in the markdown format, together with extracted links:

curl -X POST https://api.firecrawl.dev/v2/search \
  -H "Content-Type: application/json" \
  -d '{
    "query": "firecrawl web scraping",
    "limit": 3,
    "scrapeOptions": {
      "formats": ["markdown", "links"]
    }
  }'

The guide also documents a separate workflow: search first, filter or process the results, then call /scrape for selected URLs. This lets an application choose which pages to retrieve before performing the content extraction.

Which settings control the search?

The API reference defines query as the search text and sources as the selection of result types. The limit setting applies per source type when several types are requested.

includeDomains restricts results to specified domains, while excludeDomains removes specified domains. These settings accept hostnames without a protocol or path and are mutually exclusive. For a website audit, a domain restriction therefore defines which hostnames the search covers.

The reference also documents location and country for geographic targeting, recommending that both be set together. Its tbs parameter accepts predefined time periods and custom date ranges.

How can the retrieved page be checked?

Firecrawl's Scrape guide explains that a successful API request can accompany an error response from the target page. It recommends checking the API response's success value and the target page's metadata.statusCode separately. A page error can also have detail in metadata.error.

The same guide distinguishes cleaned HTML, requested as html, from unmodified HTML, requested as rawHtml. These formats provide different views of the retrieved document. For a website owner checking extraction, comparing the requested output with the page helps identify which text the application receives.

FAQ

Can I retrieve only selected search results?

Yes. Firecrawl's Search documentation describes searching first and then scraping selected result URLs. You can filter or rank the search results before retrieving their page content.

Can I request several content formats?

Yes. The Scrape documentation supports multiple output formats in a request, including Markdown, HTML and links. The output keys correspond to the formats you choose.

How many results can I request?

The Search API reference gives limit a default of 10 and a range from 1 to 100. When you request multiple source types, the limit applies to each type.

Sources

Check how a page reads to answer engines. Open the GEO score analyzer