Reorder

Endpoint: POST https://4zapi.com/v1/rerank

Given a hint, the model will return one or more predicted completions and can also return the probability of the surrogate marker for each position.

Complete creation for the provided prompts and parameters

Official documentation: https://docs.siliconflow.cn/cn/api-reference/rerank/create-rerank

Parameters

• Content-Type (header, string, required): Request body Content-Type, usually application/json
• Accept (header, string, required): Expected response Content-Type
• Authorization (header, string): Bearer API Key
• X-Forwarded-Host (header, string): Forwarded original Host (proxy scenario)

Request body (application/json)

• documents (array<string>, required): List of documents to rerank
• instruct (string, required): Rerank instruction / extra notes
• model (string, required): Model name / model ID
• query (string, required): Query text
• top_n (integer, required): Return the top N results

Request example

{
  "documents": [
    "文本排序模型广泛用于搜索引擎 and 推荐系统中，它们根据文本相关性对候选文本进行排序",
    "量子计算是计算科学的一个前沿领域",
    "预训练语言模型的发展给文本排序模型带来了新的进展"
  ],
  "instruct": "Given a web search query, retrieve relevant passages that answer the query.",
  "model": "qwen3-rerank",
  "query": "什么是文本排序模型",
  "top_n": 2
}