Skip to main content
POST
/
v1
/
models
/
{model_id}
/
predict
Predict
curl --request POST \
  --url https://api.example.com/v1/models/{model_id}/predict \
  --header 'Content-Type: application/json' \
  --data '
{
  "dataset": {
    "columns": [
      {
        "name": "<string>"
      }
    ],
    "connection_config": {
      "uri": "<string>",
      "connector_credential_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "connector_credential_name": "<string>",
      "connector_type": "s3",
      "region": "<string>",
      "file_format": "<string>"
    },
    "filter": {
      "col": "<string>",
      "kind": "cmp",
      "value": "<string>"
    }
  }
}
'
{
  "job_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "message": "<string>"
}

Headers

​
authorization
string | null
​
refresh-token
string | null

Path Parameters

​
model_id
string<uuid>
required

Body

application/json

POST /v1/models/{model_id}/predict -- Batch predictions.

​
dataset
Dataset · object
required

Inline dataset spec (connection config + columns + optional filter) describing the rows to score. The server builds a typed trainer Dataset from it and ships it to the inference worker, which materializes inside its container.

Response

Successful Response

Base response returned by async job-submission endpoints.

Keyed on job_id. Model identity (when applicable) is materialized at completion and returned inside JobResultResponse.result_data.

​
job_id
required

Server-assigned id of the submitted job

​
status
enum<string>
required
Available options:
pending,
running,
completed,
failed
​
message
string
required