Accepts a customer message or conversation and returns the predicted intent with confidence scores.
API token from your Handled dashboard
The customer message or text to classify
2048Prior turns in the conversation for context
Number of top intents to return
1 <= x <= 5Minimum confidence score to include in results
0 <= x <= 1Successful classification
Unique identifier for the inference request
The input message that was classified
Ranked list of intent predictions
The highest-confidence intent
track_order, cancel_order, return_item, refund_request, product_inquiry, change_address, payment_issue, complaint, compliment, other Server-side inference time in milliseconds
Version of the model used