Skip to main content
OpenOpen8 uses a ratio-based pricing system to control how quickly your credits are consumed. Every request deducts a calculated number of units from your quota based on the tokens used and a set of configurable multipliers.

​
How quota consumption is calculated

When a request completes, OpenOpen8 calculates the quota consumed using this formula: 
quota consumed = tokens × model_ratio × group_ratio
  • tokens — the number of tokens processed in the request (input + output, or as determined by the provider)
  • model_ratio — a multiplier specific to the model used
  • group_ratio — a multiplier applied to your account’s group
Input tokens and output tokens can carry different ratios. For models where output is significantly more expensive than input (like most frontier models), the output ratio is higher to reflect that cost difference.
Model ratios are set to reflect real upstream costs. A model that costs roughly 5× more than the baseline has a ratio of approximately 5, keeping credit balances proportional to real spending.

​
Model ratios

Each model has two ratios:
RatioWhat it controls
Input ratioMultiplier applied to input (prompt) tokens
Output ratioMultiplier applied to output (completion) tokens
For example, if gpt-4o has an input ratio of 5 and an output ratio of 15, a request that uses 1,000 input tokens and 500 output tokens would consume (1000 × 5) + (500 × 15) = 12,500 quota units.
You can view the effective pricing for each model in the OpenOpen8 dashboard. Ratios are set to closely match actual upstream provider costs.

​
Group ratios

Group ratios apply a blanket multiplier to all requests from users in a given group. This enables:
  • Discounted tiers — a group ratio below 1.0 means credits are consumed more slowly
  • Standard tiers — the default group ratio is 1
  • Premium tiers — a group ratio above 1.0 for higher-cost access
The final quota consumed is always the product of the model ratio and the group ratio. A model ratio of 5 combined with a group ratio of 0.5 results in an effective multiplier of 2.5.

​
Checking your effective rates

To understand how much a specific model costs in credits:
  1. Log in to the OpenOpen8 dashboard at openopen8.ai.
  2. Check the model pricing information available in the dashboard.
  3. Review your usage in Logs to see actual credit consumption per request.
The Logs page shows the exact credit cost of each request, broken down by input and output tokens. Use this to estimate costs for your workload before scaling up.