Search API returning source-linked web excerpts selected for agent queries.
Features
- Web search API
- Paragraph-level relevance filtering
- Source-linked selected excerpts
- API, SDK, CLI, and MCP
- Compatible with existing calls
Use Cases
- Agent web retrieval
- RAG context acquisition
- Research and fact finding
- Data enrichment and information extraction
FAQ
The new Firecrawl /search uses a custom relevance model to score paragraphs, lists, and tables in search results and returns the excerpts that best answer a query. Built for agent search, RAG, and research workflows, it is available through the API, SDKs, CLI, and MCP while preserving the existing API shape for current integrations. Core capabilities include: Web search API, Paragraph-level relevance filtering, Source-linked selected excerpts.
Common scenarios include: Agent web retrieval, RAG context acquisition, Research and fact finding.
