Developer guide

A practical web search API for your product.

Send a natural-language question, set the research controls you need, and receive structured search results, citations, streaming events, and answer telemetry.

Start with one request

Your live API base is already in the docs.

This site automatically uses the domain where it is opened. You do not need to edit localhost URLs after deployment.

Live API base

Loading…

Open API explorer →Read full docs

POST Loading…

{
  "query": "Compare vector databases for RAG",
  "depth": "deep",
  "safe": true,
  "max_output_tokens": 1200,
  "stream": false
}

Endpoints

A compact surface for search and research.

POST/searchJSON or Server-Sent Events

GET/searchQuery-string search request

GET/streamStreaming convenience route

GET/sourcesRetrieval-only source list

GET/healthDiagnostics and provider state

Output-token control

Set max_output_tokens between 128 and 20,000. Omit it when the normal depth budget is right for the request.

Streaming support

Set stream: true or use /stream to display real search and writing progress in your UI.

Need more detail?

The interactive documentation lists request fields, streaming events, depth profiles, models, and operational endpoints.

Open documentation →