Model Response
A response produced by the Responses API.
output lists the items the model produced, in order: the ReasoningItem of a thinking model comes first, then the assistant MessageItem, with FunctionCallItem and WebSearchCallItem interleaved as the model calls tools. Because the API is stateless, feeding output straight back into the next request's input is what continues the conversation.
Example:
val response = client.createResponse("What is the capital of France?")
println(response.outputText)
println(response.usage?.totalTokens)See also
Constructors
Properties
Set when status is ResponseStatus.FAILED, describing what went wrong.
The function calls in output, in the order the model emitted them.
Set when status is ResponseStatus.INCOMPLETE, saying why the response was cut short.
Items the model produced, in generation order.
Text of every OutputTextPart the model produced, concatenated in order.
Whether the response completed, was truncated, or failed.
Token usage statistics; absent while the response is still in progress.