Billing
XNova calculates cost by model, group, and request type. Text models are usually billed by token usage, while image models are billed by image tier.
Text Models
Text billing uses input, output, cache usage, and the current group multiplier.
Image Models
Image generation is charged by image size tier. For gpt-image-2, the recommended independent image prices are:
| Tier | Price per image |
|---|---|
| 1K | 0.4 |
| 2K | 0.5 |
| 4K | 0.6 |
Multiple images are charged by the actual image count.
Balance and Subscription Groups
- Balance group: charges actual request cost.
- Subscription group: text models may use a subscription multiplier.
- Independent image multiplier: when enabled, image generation uses the image price table instead of the text multiplier.