跳转到主要内容
Use this endpoint to convert text into numerical vector representations (embeddings). You can then use these vectors for semantic search, retrieval-augmented generation (RAG), clustering, and classification — any task that benefits from measuring meaning-based similarity between texts. Endpoint: POST https://api.qhaigc.net/v1/embeddings

​
Supported Models

Model IDDimensionsMax TokensLanguages
bge-m310248192100+ languages
text-embedding-3-large30728192Multilingual
Use bge-m3 for multilingual content or when you need long-context support (up to 8192 tokens per input). Use text-embedding-3-large when you need higher-dimensional embeddings for maximum retrieval precision.

​
Request Parameters

​
model
string
必填
The embedding model to use. Examples: bge-m3, text-embedding-3-large.
​
input
string | array
必填
The text to embed. Pass a single string, or an array of strings to embed multiple texts in one request. Each string must be within the model’s token limit.
​
encoding_format
string
The format of the returned embeddings. Use "float" (default) for an array of 32-bit floats, or "base64" for a base64-encoded string.

​
Response Fields

​
object
string
Always "list".
​
data
array
Array of embedding objects, one per input string. Results are returned in the same order as the inputs.
​
model
string
The model that produced the embeddings.
​
usage
object
Token usage for this request.

​
Code Examples

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-api-key-here",
    base_url="https://api.qhaigc.net/v1"
)

response = client.embeddings.create(
    model="bge-m3",
    input="Qhaigc is an AI API platform focused on low-cost exploration of AIGC capabilities."
)

embedding = response.data[0].embedding
print(f"Embedding dimensions: {len(embedding)}")
print(f"First 5 values: {embedding[:5]}")

​
Batch embedding multiple texts

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-api-key-here",
    base_url="https://api.qhaigc.net/v1"
)

texts = [
    "Semantic search finds documents by meaning, not keywords.",
    "RAG combines retrieval with language model generation.",
    "Vector databases store embeddings for fast similarity search."
]

response = client.embeddings.create(model="bge-m3", input=texts)

for item in response.data:
    print(f"Text {item.index}: {len(item.embedding)}-dimensional vector")

​
Example Response

{
  "object": "list",
  "data": [
    {
      "embedding": [
        -0.06332587,
        -0.019955011,
        -0.03361425
      ],
      "index": 0,
      "object": "embedding"
    }
  ],
  "model": "bge-m3",
  "usage": {
    "prompt_tokens": 21,
    "total_tokens": 21
  }
}

​
Common Use Cases

Build a RAG pipeline by embedding your knowledge base, storing vectors in a vector database, retrieving the top-k relevant chunks at query time, and passing them as context to a chat model.
Detect near-duplicate content or find semantically similar items in your dataset by comparing embedding vectors.
Use embeddings as features for downstream classifiers, or group semantically similar texts with k-means or hierarchical clustering.