LightRAG Query icon

LightRAG Query

Query documents via LightRAG

LightRAG Query

AI-generated

Summary

Send a query string to the LightRAG API to retrieve a generated response based on selected modes and response formats.

Inputs

  • Query (required) — The question string to send to the LightRAG API for retrieval and response generation.
  • Mode (required) — The querying mode to use; options include Mix, Naive, Global, Hybrid, Local, and Bypass, which determine how the query is processed.
  • Response Type (required) — Determines the format of the response, selectable as Multiple Paragraphs, Single Paragraphs, or Bullet Points.
  • Options — Additional optional parameters to customize the query, such as Top K nodes, chunk size, maximum token limits, and whether to enable reranking.

Output shape

a JSON array containing the API response object(s) from the LightRAG service

The response is returned as array-wrapped JSON data as received from the LightRAG API, containing generated answers based on the query and options provided.

Discussion