O-mega uses a straightforward, step-based credit system that makes it easy to understand and predict your usage. Rather than charging based on the complexity of tasks or the specific AI model you use, credits are tied directly to the actions your agents take—giving you full transparency into how your balance is consumed.
How Credits Work
Every time your agent takes a step during a browser session or computer session, credits are deducted from your account. The system is designed to be simple: one step costs one credit, regardless of which AI model powers your agent or how computationally intensive the underlying task might be.
This means you don't need to worry about unexpected cost increases when using more advanced models. Whether your agent is powered by a fast, lightweight model or a sophisticated reasoning model, the credit cost remains consistent based purely on the number of steps taken.
A key implication of this design is that switching between AI models does not affect your credit consumption. A 10-step browser automation uses exactly 10 credits whether it runs on the default model or a premium alternative. Similarly, a computer session that generates a complete website in a single step uses just 1 credit.
Steps vs Actions
Understanding the difference between steps and actions is important for making the most of your credits. A step is not necessarily a single discrete action—it's a logical unit of work that can encompass multiple related actions performed together.
Your agent intelligently groups related actions into a single step to maximize efficiency. For example, navigating to a webpage and extracting data from it might happen within the same step. Clicking a button and filling out a form field can be combined. When generating code in a computer session, an entire file can be produced as one step.
Common step combinations include:
- Navigate to a page AND extract data in one step
- Click a button AND fill out a form field together
- Generate an entire code file in a single computer step
- Take a screenshot while performing other tasks
This batching behavior means your agents work efficiently without you needing to micro-manage credit usage.
Credit Balance
Your credit balance is tied to your subscription plan. Each plan includes a monthly allocation of credits that automatically resets at the start of your billing cycle. The amount included varies by plan tier—higher tiers include more credits to support larger workloads.
You can monitor your remaining credits at any time in your account settings. The balance updates in real-time as your agents work, so you always know exactly where you stand.
Credit Overage
If you use all your monthly credits before your billing cycle resets, you have flexibility in how to proceed. O-mega offers two paths forward depending on your preferences.
The first option is to upgrade your plan to one with a higher monthly credit allocation. This is ideal if you consistently find yourself needing more capacity.
Alternatively, you can enable credit overage, which allows your agents to continue working beyond your included allocation. When overage is enabled, additional usage is tracked and billed at the end of your billing cycle at a per-credit rate.
You control whether overage is enabled in your subscription settings. If you prefer to stay strictly within your plan limits, you can leave overage disabled—in which case agent automation will pause when credits run out and resume automatically when your next billing cycle begins.
Tips for Efficient Credit Usage
Since credits are consumed step-by-step, there are practical ways to optimize your usage without sacrificing what your agents can accomplish.
Writing clear, specific prompts helps your agents complete tasks in fewer steps. When your agent understands exactly what you need, it can work more directly without exploratory steps.
Breaking complex tasks into focused sub-tasks prevents unnecessary exploration. Instead of giving your agent a vague directive, be specific about what you want accomplished.
Using task templates is another efficiency lever. Pre-defined workflows are already optimized for common use cases and tend to execute more predictably than ad-hoc instructions.
Related: Browser Sessions Overview | Computer Sessions Overview