上下文相关的块嵌入模型

正式环境

https://api.302.ai

POST

/voyageai/v1/contextualizedembeddings

Voyage 的上下文相关块嵌入端点除了接受查询和完整文档外，还接受文档块，并返回包含上下文相关块向量嵌入的响应。这些上下文相关块嵌入不仅捕获每个块内的局部细节，还捕获来自整个文档的全局、粗粒度元数据。

价格：0.0002 PTC/1000 token

请求参数

Authorization

在 Header 添加参数

Authorization

，其值为在 Bearer 之后拼接 Token

示例：

Authorization: Bearer ********************

Header 参数

Body 参数application/json必填

示例

返回响应

🟢200成功

application/json

Body

请求示例

Shell

JavaScript

Java

Swift

curl --location --request POST 'https://api.302.ai/voyageai/v1/contextualizedembeddings' \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data-raw '{
    "inputs": [
      ["doc_1_chunk_1", "doc_1_chunk_2"],
      ["doc_2_chunk_1", "doc_2_chunk_2"]
    ],
    "input_type": "document",
    "model": "voyage-context-3"
  }'

响应示例

{
  "object": "list",
  "data": [
    {
      "object": "list",
      "data": [
        {
          "object": "embedding",
          "embedding": [0.040455919, -0.017057089, -0.007927173, "...", 0.042205364],
          "index": 0
        },
        {
          "object": "embedding",
          "embedding": [0.03946501, -0.015512259, -0.009809222, "...", 0.033990096],
          "index": 1
        }
      ],
      "index": 0
    },
    {
      "object": "list",
      "data": [
        {
          "object": "embedding",
          "embedding": [0.047289684, -0.042560715, -7.7669e-05, "...", 0.038691558],
          "index": 0
        },
        {
          "object": "embedding",
          "embedding": [0.04198277, -0.036370102, -0.00352195, "...", 0.031879965],
          "index": 1
        }
      ],
      "index": 1
    },
  ],
  "model": "voyage-context-3",
  "usage": {
    "total_tokens": 24 
  }
}

修改于 2025-11-04 10:34:44

Rerank（重排序）

Scripts（生成视频素材文案）