AI Units vs Modern Activities

Does using Modern Activities with Document Understanding increase AI Units usage & Do retries or exception handling increase AI Units consumption?

AI Units are consumed when Modern UI Automation uses:

    Computer Vision (CV) with AI-based detection
    Screen Understanding (AI-driven UI detection)
    AI-powered selectors fallback
    Autopilot or GenAI suggestions

Screen Understanding uses:

    AI models
    OCR
    Vision-based classification

:check_mark: Best practices:

    Prefer selectors over Computer Vision
    Use one Use Application/Browser container
    Disable CV unless required
    Avoid repeated Screen Understanding calls
    Cache extracted results
    Use OCR sparingly

Normal selector-based automation does not consume AI Units.

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.