Skip to main content
Poixe AI’s model prices and discount information are all shown on the pricing page:

Model Pricing and List

View supported models, standard prices, billing units, and the latest changes.

User Tiers and Benefits

View your tier, rate limits, discounts, and free capacity.
Poixe AI’s model costs are not fixed and are affected by several factors. You can understand them with the following logic:

​
User level (Level) affects discounts and available resources

New users start at Level 0. As your cumulative top-ups and usage grow, your level is raised automatically. A higher level usually means:
  • Bigger discounts: the same model costs less
  • More capacity: higher rate limits, concurrency, and so on
Levels are assigned automatically by the system based on cumulative top-ups, and upgrades take effect immediately.

​
Cost per request = standard billing × discount

The cost of a single request is usually calculated using the model’s official standard billing method (for example, by input/output tokens), and then your Level discount is applied to produce the final charge.

​
The same model may have multiple providers, with different standard pricing

The same model name (for example gpt-5.2) may be offered by multiple model providers on the platform. The standard pricing and response structure can differ between providers.
  • No provider specified: uses the default route
    • The system balances requests across multiple providers to improve capacity and availability
    • The trade-off: response details may differ slightly between providers (refer to the actual response)
  • Provider specified: for example openai
    • Billing and responses follow that provider’s standard pricing and response format

​
Top-up currency

Top-ups on the platform are priced and settled in US dollars (USD).
To quickly check how much a given model will actually cost you, the most direct way is to: