Skip to main content
When setting up how you charge for your agent, you can choose from four different billing methods. Each method determines how costs are calculated and applied for every run.
For all methods, be sure to set margins that cover any additional resources or overhead you spend outside of MuleRun. 5% of the revenue will be retained as platform fee, with the remaining 95% allocated to you.

​
1. Per Minute

  • How it works: You set a unit price per minute. Users are billed based on how long the agent runs.
  • Example: If your rate is $0.10/minute and the agent runs for 7 minutes, the total cost is $0.70.
  • Best for: Workflows where runtime is the main factor in cost (e.g., data processing, monitoring tasks).

​
2. Per Step

  • How it works: You charge for each step in the workflow. A “step” refers to actions in the execution path that runs LLM models.
  • Example: If your rate is $0.05/step and a workflow runs 20 steps, the total cost is $1.00.
  • Best for: Workflows where the number of steps, not time, best reflects cost.

​
3. Per Run

  • How it works: You charge a fixed price for each run, regardless of runtime or step count.
  • Example: If your rate is $2.00/run, the workflow always costs $2, whether it takes 30 seconds or 30 minutes.
  • Best for: Simple, predictable pricing when every run has roughly the same value.

​
4. Cost-Plus

  • How it works: You charge a margin on top of the actual cost of resources used on MuleRun that Mulerun calculates(the base cost on MuleRun).
    You can set your margin in two ways:
    • By Fixed Fee: Add a set dollar amount to the actual cost.
      • Example: If actual cost = $1.50 and your fee = $0.50, user pays $2.00.
    • By Fixed Percentage: Add a percentage of the actual cost.
      • Example: If actual cost = $1.50 and margin = 20%, user pays $1.80.
  • Best for: When your costs vary widely and you want to ensure coverage while keeping pricing transparent.

​
5. Creator Metering

  • How it works: This method allows you to set your own pricing dynamically through the Metering API. Your system reports billing data to MuleRun at defined intervals, determining how much customers are charged based on your defined metrics.
  • Example: If your workflow processes data in batches, you can report cost per batch, per token, or any metric you define through your Metering API logic.
  • Best for: Advanced creators who want full control over how their billing is calculated or need flexible, usage-based pricing models.
  • Development guide: For setup and implementation details, follow the Metering Report API Specs.
⌘I