Pricing
Transparent pay-as-you-go pricing across all models
AI Gateway applies one unified billing scheme across providers:
- Per-token billing — input and output tokens are priced separately. Prompt-cache hits settle at the cache read / cache write rate.
- Per-call billing — image and other discrete models are billed by request count, regardless of parameters.
- No monthly fee, no minimums — pay only for what you use; every model settles in a single account.
- Currency — prices are quoted in CNY (¥); the live rate published on the server is authoritative.
At a glance
| Type | Billing | Example |
|---|---|---|
| Chat | per-token | claude-sonnet-4-6: ¥1.20 in / ¥6.00 out per 1M tokens |
| Codex-style | per-token | gpt-5.3-codex: ¥0.088 in / ¥0.70 out per 1M tokens |
| Image | per-call | gpt-image-2: ¥0.10 per call |
For the full price list and current model tiers, see the Models page or the live Model Marketplace.