AI Action (Media Analysis)

The AI Media Understanding node allows the bot to analyze media (images or videos) and generate AI-based insights such as summaries, descriptions, or observations using an AI model.


Screenshot 2026-02-02 at 5.24.13 PM.png

Screenshot 2026-02-02 at 5.25.07 PM.png

Screenshot 2026-02-02 at 5.26.23 PM.png
#

How It Works

  1. The user provides a media attachment or media URL.

  2. The AI Media Understanding node sends the media along with the configured prompt to the AI model.

  3. The node returns the AI-generated response on Success or triggers failure in case of an error.


#

Configuration Steps

  1. Add the AI Media Understanding node to the flow.

  2. Enter a Prompt (example: Summarize the media in 3 sentences).

  3. Provide the Media URL (static or via variable).

  4. Select the AI Model based on your convenience (e.g., gemini-2.0-flash).

  5. Add a valid Google Gemini API Key.

  6. Connect the Success and Fail paths to the next nodes.


#

Supported Media

  • Images (JPG, PNG, etc.)

  • Videos (maximum duration: 60 minutes)


#

Common Failure Reasons

  • Media URL is missing or not publicly accessible

  • Invalid or missing API key

  • Unsupported media format

  • Video exceeds 60-minute limit


#

Best Practices

  • Ensure media URLs are publicly accessible

  • Use clear and specific prompts