Classification - Beta

Classify text into one or more predefined categories using LLM structured outputs.

POST /api/ai/v1/classification

cURL

Support Ticket Priority (Single-Label)

curl --request POST \
  --url https://{subdomain}.domo.com/api/ai/v1/classification \
  --header 'Content-Type: application/json' \
  --header 'X-DOMO-Developer-Token: <api-key>' \
  --data @- <<EOF
{
  "input": [
    {
      "type": "text",
      "text": "URGENT: Our production dashboard is down and we can't access any of our KPI data. This is affecting our entire sales team and we need this resolved immediately!"
    }
  ],
  "labels": [
    {
      "label": "critical",
      "description": "System outage or data loss affecting multiple users or business operations"
    },
    {
      "label": "high",
      "description": "Significant functionality issue affecting user productivity"
    },
    {
      "label": "medium",
      "description": "Feature not working as expected but workarounds available"
    },
    {
      "label": "low",
      "description": "Minor issue, cosmetic bug, or feature request"
    }
  ],
  "classificationType": "SINGLE_LABEL"
}
EOF

200

{
  "output": [
    "critical"
  ],
  "classificationType": "SINGLE_LABEL",
  "model": "domo.domo_ai.domogpt-medium-v1.2:anthropic",
  "isCustomerModel": false,
  "sessionId": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "requestId": "f9e8d7c6-b5a4-3210-fedc-ba9876543210"
}

Authorizations

X-DOMO-Developer-Token · string · header · required

Body

application/json

Request for classifying content into predefined categories using the Messages API.

The Classification service uses LLM structured outputs to reliably classify content into one or more predefined categories. The service dynamically generates JSON schemas based on the provided labels to ensure type-safe responses.

input · object[] · required

The input message content to classify (text, images, documents, etc.)

Text-based message content.

Option 1

Option 2

Option 3

Show child attributes

labels · object[] · required

The list of possible classification labels with optional descriptions (2-50 labels required)

Show child attributes

classificationType · enum<string> · required

The type of classification: SINGLE_LABEL or MULTI_LABEL

Available options: SINGLE_LABEL, MULTI_LABEL

instructions · string

Optional instructions for classification behavior

model · string

The ID of the model to use (optional, uses default if not specified)

modelConfiguration · object

Additional model parameters (temperature, max_tokens, etc.)

Show child attributes

sessionId · string<uuid>

AI session ID for associating with existing session

reasoningConfig · object

Configuration for reasoning behavior

Show child attributes

Response

200

application/json

Successful classification response.

Response for text classification. Supports both single-label and multi-label classifications uniformly. All classification responses return a list of labels via the output() method.

output · string[]

The classification output - a list of selected labels.

classificationType · enum<string>

The type of classification (SINGLE_LABEL or MULTI_LABEL).

Available options: SINGLE_LABEL, MULTI_LABEL

model · string

The ID of the model used for classification.

isCustomerModel · boolean

Whether the model is a customer model (vs Domo provided).

sessionId · string<uuid>

The AI session ID associated with this response.

requestId · string<uuid>

The request ID associated with this response.

modelProviderUsage · object

The token usage from the model provider.

Show child attributes