POST https://api.wylon.cn/v1/embeddings

Embeddings

Convert text into fixed-length vector representations for semantic search, similarity, clustering, RAG, and other downstream tasks.

Authorization

Authorizationstring · headerrequired

Bearer token. Create one in the dashboard.

Request body

modelstringrequired

Embedding model ID, e.g. BAAI/bge-large-zh, Qwen/Qwen3-Embedding. Full list in the model catalog.

inputstring · string[]required

Text to embed. Pass a single string or an array (up to 64 per request). Each item must fit within the model's context length.

encoding_formatenum"float"

Output format: "float" returns JSON float arrays; "base64" returns a compact Base64 byte string, useful for reducing payload size.

userstringoptional

End-user identifier for abuse monitoring.

Response

objectstring

Always list.

modelstring

The embedding model ID that actually served this request.

dataarray

One entry per input item, in the same order as input.

objectstring

Always embedding.

indexinteger

Index in the input array.

embeddingnumber[] · string

The vector itself; with encoding_format=base64 this is a Base64 string that you decode to a float array.

usageobject

Only prompt_tokens and total_tokens are reported (no generation tokens).

Example response

{
  "object": "list",
  "model": "BAAI/bge-large-zh",
  "data": [
    {
      "object": "embedding",
      "index": 0,
      "embedding": [-0.012, 0.034, 0.281, "…"]
    }
  ],
  "usage": { "prompt_tokens": 12, "total_tokens": 12 }
}

Bad request, authentication failure, or an exceeded rate limit.

Example — 400 input too long

{
  "error": {
    "type": "invalid_request_error",
    "message": "Input text exceeds model context length (max 8192 tokens)."
  }
}