XuanShu API

Embeddings API

OpenAI-compatible Embeddings endpoint and capability boundaries.

Last verified: 2026-07-23

Connection and authentication

Unified OpenAI-compatible Base URL: https://www.xuanshuapi.com/v1

OpenAI-compatible APIs use Authorization: Bearer <YOUR_API_KEY>; Claude Messages also accepts x-api-key: <YOUR_API_KEY>; Gemini v1beta uses x-goog-api-key: <YOUR_API_KEY>.

Embeddings API

POST /v1/embeddings is available only for models enabled for vector requests in the current configuration.

Embeddings are available only for models enabled in the current configuration.

Request and call

Embeddings accepts a model and a string or string-array input.

curl https://www.xuanshuapi.com/v1/embeddings \
  -H "Authorization: Bearer <YOUR_API_KEY>" -H "content-type: application/json" \
  -d '{"model":"<MODEL_FROM_MODELS>","input":["first document","second document"]}'

Response and streaming behavior

The data array returns vectors and indexes in input order; this endpoint is not streaming.

HTTP 200
{"object":"list","data":[{"object":"embedding","index":0,"embedding":[0.012,-0.034]},{"object":"embedding","index":1,"embedding":[0.056,0.078]}],"usage":{"prompt_tokens":4,"total_tokens":4}}

Key parameters

Parameter	Description
model	Must be enabled for vector requests.
input	A string or array of strings.
encoding_format	Set only when supported by the target model.

Typical failures

For 400 inspect the input type; for 404 inspect the model ID; models without vector support fail.

HTTP 400  # invalid input type
HTTP 404  # model not found or unavailable
HTTP 429/5xx  # bounded retry candidate

Next steps

Available models, multipliers, quotas, and some advanced capabilities depend on the current key and console configuration.