Embeddings / Alibaba Cloud

Text Embeddings API

Convert text into high-dimensional embedding representations for semantic search, text clustering, recommendation systems, and other scenarios. The API uses synchronous invocation, returning embedding results immediately after the request, with no polling needed.

Request Endpoint

POSThttps://nexusflow.vip/v1/embeddings

Supported Models

Model ID	Provider	Embedding Dimensions	Price
`text-embedding-v4`	Alibaba Cloud	2048/1536/1024/768/512/256/128/64	¥0.5/million tokens

Request Parameters

Parameters	Type	Required	Description
`model`	string	*	Embedding model ID, fixed value: text-embedding-v4.
`input`	string \| string[]	*	Text to be embedded. Can be a single string or a string array (batch processing).
`encoding_format`	string	-	Embedding encoding format in the response. Options: "float" (default) or "base64".

Code Examples

curl -X POST 'https://nexusflow.vip/v1/embeddings' \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text-embedding-v4",
    "input": "NexusFlow is a unified large model API platform"
  }'

Response Format

{
  "object": "list",
  "data": [
    {
      "object": "embedding",
      "index": 0,
      "embedding": [0.0023064255, -0.009327292, 0.015797347, ...]
    }
  ],
  "model": "text-embedding-v4",
  "usage": {
    "prompt_tokens": 12,
    "total_tokens": 12
  }
}

Response Fields

Field	Type	Description
`object`	string	Fixed value "list", indicating the response is an embedding list.
`data[].embedding`	array	Float embedding array; dimensions depend on the model.
`data[].index`	integer	Index position corresponding to the input text (starting from 0).
`model`	string	The model ID used for this request.
`usage`	object	Token usage statistics, including prompt_tokens and total_tokens.

Application Scenarios

Semantic Search

Convert queries and documents into embeddings, then perform semantic-level search via embedding similarity, which is more precise than keyword matching.

Text Clustering

Perform clustering analysis on large volumes of text, automatically discovering implicit topics and patterns, suitable for sentiment analysis and content categorization.

Recommendation Systems

Recommend related articles, products, or services based on content embedding similarity, improving personalized experience.

Anomaly Detection

Identify anomalous content or behavior by calculating the distance between text embeddings and normal samples, detecting deviations from normal patterns.

Chat Completions API

View text generation API documentation

Error Code Reference

View error codes and handling methods

Full Pricing

View pricing for all models