AI Agent connector Restriction

Camunda 8 Topics
1

When I try to call AI Agent connector many times, I’ received an Error message in Operate:

“Maximum number of model calls reached (modelCalls: 10, limit: 10)“.

Is it possible to avoid this Restriction ? If yes, then how.

Latest “connectors bundle” works in Docker container with minimum default params. Test diagram is very simple:

Connector_Restriction_10calls-2
Connector_Restriction_10calls-2832×409 15.6 KB

2

Hi @volodya327,

Yes, you can definitely avoid this restriction! The “Maximum number of model calls reached” error occurs because the AI Agent connector has a default limit of 10 model calls, but this is configurable.

How to Increase the Limit

The Maximum model calls setting can be adjusted in the connector configuration:

  1. Open your BPMN in Web Modeler or Camunda Modeler
  2. Select your AI Agent element (either the AI Agent Sub-process or AI Agent service task)
  3. Go to the Properties panel and find the Limits section
  4. Set “Maximum model calls” to your desired number (e.g., 20, 50, 100, etc.)
  5. Save and redeploy your process

After redeployment, the agent will use your configured value instead of the default 10.

Why This Limit Exists

This limit is a safety mechanism to prevent runaway AI agents from making excessive API calls, which could:

  • Consume your AI model quota/credits quickly
  • Create infinite loops in agent reasoning
  • Impact performance

Additional Resources

Let me know if you need help with the configuration or if you have any other questions!

3

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.