Embeddings¶

Convert text or multimodal content to vectors for later retrieval or similarity calculations.

POST https://token.moi.matrixorigin.cn/v1/embeddings

Before you call¶

Prepare credentials with Genesis access and a model ID that supports Embeddings.

Request body¶

Replace $GENESIS_ACCESS_TOKEN with an access token or service-account API key, and $MODEL_ID with the model ID to call.

Request example

curl -X POST \
  "https://token.moi.matrixorigin.cn/v1/embeddings" \
  -H "Authorization: Bearer $GENESIS_ACCESS_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "'"$MODEL_ID"'",
    "input": [
      "MatrixOne is a database for the intelligent era.",
      "Genesis provides model invocation APIs."
    ],
    "encoding_format": "float"
  }'

Parameter description

Field	Type	Required	Description
`model`	string	Yes	Embeddings model ID to call.
`input`	string or array of string or array of object	Yes	Text to convert to vectors.
`encoding_format`	string	No	Vector encoding format. The example uses `float`.
`dimensions`	integer	No	Expected vector dimensions; use only when the selected model supports them.

Successful response¶

The service returns one corresponding vector result for each input.

Response example

{
  "object": "list",
  "model": "<MODEL_ID>",
  "data": [
    { "object": "embedding", "index": 0, "embedding": [0.018, -0.042, 0.337] }
  ]
}

Field description

Common fields

Field	Type	Description
`object`	string	Response object type; the example is `list`.
`model`	string	Model ID that processed the request.
`data`	array of object	Vector results for this input; each item corresponds to one input text.

Result item

The following table expands each item in the example data array.

Field	Type	Description
`object`	string	Result object type; the example is `embedding`.
`index`	integer	Position corresponding to the `input` array.
`embedding`	array of number or string	Vector for this text; the actual type depends on the requested encoding format.

Error response¶