> ## Documentation Index
> Fetch the complete documentation index at: https://docs.crazyrouter.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Rerank

> Semantically rerank search results

> Дата обновления: 2026-06-06

## Overview

Reorder a set of documents by semantic relevance to a query. Commonly used as the second-stage reranking step in RAG (Retrieval-Augmented Generation) pipelines.

<Note>
  Implemented following the [SiliconFlow Rerank API](https://docs.siliconflow.cn/cn/api-reference/rerank/create-rerank) format.
</Note>

## Supported Models

| Model           | Description                               |
| --------------- | ----------------------------------------- |
| `gte-rerank-v2` | Multilingual reranking model, recommended |
| `gte-rerank-v2` | Primarily English                         |

## Request Parameters

<ParamField body="model" type="string" required>
  Reranking model name, e.g. `gte-rerank-v2`
</ParamField>

<ParamField body="query" type="string" required>
  Query text
</ParamField>

<ParamField body="documents" type="string[]" required>
  List of documents to rerank
</ParamField>

<ParamField body="top_n" type="integer">
  Return the top N results. Defaults to returning all
</ParamField>

<ParamField body="return_documents" type="boolean" default="true">
  Whether to include the original document text in the response
</ParamField>

## Response Format

```json theme={null}
{
  "model": "gte-rerank-v2",
  "results": [
    {
      "index": 2,
      "relevance_score": 0.9875,
      "document": { "text": "Most relevant document content" }
    },
    {
      "index": 0,
      "relevance_score": 0.7432,
      "document": { "text": "Second most relevant document content" }
    },
    {
      "index": 1,
      "relevance_score": 0.1205,
      "document": { "text": "Less relevant document content" }
    }
  ],
  "usage": {
    "total_tokens": 128
  }
}
```

## Code Examples

<CodeGroup>
  ```python Python theme={null}
  import requests

  response = requests.post(
      "https://api.crazyrouter.com/v1/rerank",
      headers={
          "Authorization": "Bearer sk-xxx",
          "Content-Type": "application/json"
      },
      json={
          "model": "gte-rerank-v2",
          "query": "What is a vector database",
          "documents": [
              "A vector database is a database system specialized for storing and retrieving high-dimensional vectors",
              "Relational databases use tables to store structured data",
              "Vector databases support approximate nearest neighbor search, suitable for semantic retrieval scenarios",
              "Redis is an in-memory key-value store"
          ],
          "top_n": 2,
          "return_documents": True
      }
  )

  data = response.json()
  for result in data["results"]:
      print(f"[{result['relevance_score']:.4f}] {result['document']['text']}")
  ```

  ```bash cURL theme={null}
  curl -X POST https://api.crazyrouter.com/v1/rerank \
    -H "Authorization: Bearer sk-xxx" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "gte-rerank-v2",
      "query": "What is a vector database",
      "documents": [
        "A vector database is a database system specialized for storing and retrieving high-dimensional vectors",
        "Relational databases use tables to store structured data"
      ],
      "top_n": 2
    }'
  ```
</CodeGroup>

## Typical RAG Pipeline

```
User Query → Embedding Retrieval Top-K → Rerank → LLM Generates Answer
```

<Warning>
  The number of input documents for reranking should not exceed 100. Too many documents will increase latency and cost.
</Warning>