Streaming Output
Agent.runStream returns a Stream<AgentResponse>, allowing you to process
output as it is generated by the LLM in real time. This is useful for displaying
partial results to users or for handling long responses efficiently.
import 'dart:io';
import 'package:dartantic_ai/dartantic_ai.dart';
void main() async {
final agent = Agent('openai:gpt-4o');
final stream = agent.runStream('Tell me a short story about a brave robot.');
await for (final response in stream) {
stdout.write(response.output); // Output: Once upon a time, there was a brave robot named... (streaming in real-time)
}
}