Search-enabled inference

Use a Responses-capable model with hosted search when a task needs public information beyond static model recall. B2B LLM does not expose a separate /v1/web-search operation. Tool availability depends on the selected model and route.

Request

curl https://api.b2b-llm.com/v1/responses \
  -H "Authorization: Bearer $B2B_API_KEY" -H 'Content-Type: application/json' \
  -d '{"model":"gpt-5.4","input":[{"role":"user","content":[{"type":"input_text","text":"Find the official changelog for the software named in this request and summarize the relevant release."}]}],"tools":[{"type":"web_search"}],"tool_choice":"required","include":["web_search_call.action.sources"],"stream":false}'

Response

{"object":"response","status":"completed","output":[{"type":"web_search_call","status":"completed","action":{"sources":[{"type":"url","url":"https://publisher.example/changelog"}]}},{"type":"message","content":[{"type":"output_text","text":"A summary supported by the retrieved material."}]}]}

The source address is illustrative. Search metadata records provenance, not a guarantee that the generated summary correctly interpreted each source. Keep sources with the answer when your application requires an audit trail.

Making a tool available and requiring its use are different choices. Use tool_choice: required only when the task actually needs live research. For parsing JSON versus event streams and limiting unnecessary parameters, continue to search client handling.