Liner icon

Liner

Search the web and generate source-backed answers with Liner

Actions7

Answer → AI Search

AI-generated

Summary

Generate a source-backed AI answer by submitting a user question or a full conversation history to Liner's AI Search API, optionally specifying search mode, language, and model.

Inputs

  • inputMode (required) — Choose between submitting a single question string or a full JSON-formatted conversation history of messages with roles 'user' or 'assistant'. The last message must have role 'user'.
  • question (required) — When 'inputMode' is 'question', provide the user question string to send as the final message.
  • messagesJson (required) — When 'inputMode' is 'messagesJson', provide a JSON array of conversation messages with 'role' and 'content' fields.
  • model — Specify an optional AI model identifier; defaults to Liner's default model if left empty.
  • mode (required) — Select the search mode to use, either 'general' for web search or 'scholar' for scholarly search.
  • lang — Optionally specify a language code (e.g., 'en', 'ko', 'ja') to localize the answer.
  • includeEvents — Choose whether to include raw Server-Sent Event envelopes in the output for additional event data.
  • requestId — Optional client-supplied ID to be echoed back by Liner for request tracking.

Output shape

a single JSON object containing the AI-generated answer text along with source references, reasoning, and optionally raw event data if enabled.

The output includes aggregated streaming data from Liner's AI Search, processed into fields such as text, reasoning, references, and search steps. If 'includeEvents' is true, raw SSE events are also returned. Errors during streaming or API calls will throw node errors unless 'Continue On Fail' is enabled.

Links

Discussion