NEURA | IANUSTEC AI icon

NEURA | IANUSTEC AI

Interact with NEURA | IANUSTEC AI services using OpenAI-compatible API

Actions6

Rerank → Rerank

AI-generated

Summary

Rerank documents by relevance to a query using a specified model.

Inputs

  • Query (required) — The query string to rank documents against.
  • Documents (required) — A JSON array of document strings to rerank.
  • Model — The reranker model to use (e.g., rerank-multilingual-v3.0).
  • Top N — Maximum number of results to return; leave empty to return all.

Output shape

A JSON object containing the query and an array of results. Each result includes the index, relevance score, and the document content.

Results are ordered by relevance score. The operation supports advanced options to store the raw API response in the output JSON or as a binary file.

Examples

Example 1: Rank a list of news articles by relevance to a user query.

Query: 'Latest AI developments'; Documents: JSON array of article titles; Top N: 5

Example 2: Rerank search results for a multilingual query.

Query: 'machine learning'; Documents: JSON array of search snippets; Model: rerank-multilingual-v3.0

Links

Discussion