Rerank API Guide

The Turing Platform's Rerank model ranks a set of documents by relevance to a query, and is widely used in RAG (Retrieval-Augmented Generation), search engine result optimization, and similar scenarios.

Notes:

• The Rerank API is compatible with the Cohere Rerank format
• Supports 100+ languages including Chinese, English, Spanish, French, Portuguese, Japanese, Korean, German, Russian, and Indonesian
• Supports controlling the number of returned documents via the top_n parameter
• Results are returned sorted by relevance score in descending order
• For available models, see Model List

Request Parameters

Parameter	Type	Required	Description
model	string	Yes	Model ID, e.g. qwen3-rerank
query	string	Yes	Search query string
documents	list[string]	Yes	List of documents to rank
top_n	int	No	Return only the top N ranked documents; returns all by default
return_documents	bool	No	Whether to include document content in the results
rank_fields	list[string]	No	Fields to use for ranking (used when documents are dicts)
max_chunks_per_doc	int	No	Maximum number of chunks per document
max_tokens_per_doc	int	No	Maximum number of tokens per document

Response Format

Field	Type	Description
id	string	Request ID
results	list	List of ranked results
results[].index	int	Index of the document in the original list
results[].relevance_score	float	Relevance score of the document
results[].document	object	Document content (returned when return_documents=true)
meta	object	Metadata (includes token usage, etc.)

CURL

curl $TURING_BASE_URL/rerank \
  -H "Authorization: Bearer $TURING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3-rerank",
    "query": "What is a text reranking model",
    "documents": [
      "Text reranking models are widely used in search engines and recommendation systems, ranking candidate texts by textual relevance",
      "Quantum computing is a cutting-edge field of computational science",
      "Advances in pre-trained language models have brought new developments to text reranking models"
    ],
    "top_n": 2
  }'

Response Example

{
  "id": "rerank-xxxxx",
  "results": [
    {
      "index": 0,
      "relevance_score": 0.98,
      "document": {
        "text": "Text reranking models are widely used in search engines and recommendation systems, ranking candidate texts by textual relevance"
      }
    },
    {
      "index": 2,
      "relevance_score": 0.85,
      "document": {
        "text": "Advances in pre-trained language models have brought new developments to text reranking models"
      }
    }
  ],
  "meta": {
    "billed_units": {
      "total_tokens": 128
    }
  }
}

See also

• API Reference → Rerank — Complete request/response schema with Try-It and multi-language examples (cURL / Python / Node.js / Java)
• Model List (Rerank) — Available models and pricing