v1

latestOpenAPI 3.1.02026-07-2219165126.1 KB

Embeddings

Create embeddings

Convert one or more text inputs into vector embeddings.

post/v1/embeddings

Request body

modelstring required

Embedding model identifier.

dimensionsinteger

Optional target embedding dimension when supported by the model.

encoding_format'float' | 'base64'

Return vectors as JSON floats or base64-encoded binary data.

Example request

{
  "encoding_format": "float",
  "input": [
    "semantic search",
    "vector database"
  ],
  "model": "text-embedding-3-small"
}

Response

Successful Response

object'list'

modelstring required

Example response

{
  "data": [
    {
      "embedding": [
        0.0123,
        -0.0034,
        0.0912
      ],
      "index": 0,
      "object": "embedding"
    },
    {
      "embedding": [
        0.0187,
        0.0041,
        0.0844
      ],
      "index": 1,
      "object": "embedding"
    }
  ],
  "model": "text-embedding-3-small",
  "object": "list",
  "usage": {
    "prompt_tokens": 6,
    "total_tokens": 6
  }
}