Skip to main content
Relevance AI Docs
Log InSign Up
Getting StartedBuild with Relevance AIIntegrationsAccess & AdministrationEnterpriseGuides

Understanding and Managing Your Agent Credit Usage

What Affects Your Credit Usage

AI Model Selection

Different AI models have different “appetites” for credits:

  • More powerful models generally use more credits
  • Less complex tasks can often use simpler, more credit-efficient models
  • You’ll always see the exact credit cost for each model in the model selector

Knowledge Operations

When your agent needs to access or update its knowledge base:

  • Knowledge sync operations cost the same as running a tool
  • These operations power the Retrieval-Augmented Generation (RAG) capabilities that make your agents smarter

Bulk Scheduling Operations

Running your Agent on a set of values from a Knowledge table can become costly really fast. We highly recommend you calculate a rough estimate based on how many rows of data you’re running your agent on.

Best Practices for Credit Management

  • Pause tasks if unsure of total usage over a period of time
  • Match the AI model to the task complexity
  • Use your own API keys when possible for external integrations
  • Monitor which agents and tools consume the most credits
  • Consider upgrading your plan if you need more credit-efficient operations