Skip to main content
Put a Parallel block inside a Loop to run several research steps at once on each iteration. The end condition checks the combined output and decides whether the loop needs another pass. When to use: When you need iterative quality improvement with parallel task execution in each iteration. Ideal for comprehensive research workflows where multiple independent tasks contribute to overall quality, and you need to repeat until quality thresholds are met. Install dependencies:
loop_with_parallel_steps_stream.py