How billing works
查看 API 实际采用的积分成本和计费单位。
- 显示的金额是已发布的 API 费率。已计费但失败的操作将获得退款。
- 已发布 API 费率
- 计费单位
Rate card
ServicePriceHow it is applied
Gemini 3.1 Pro Preview500 in · 3,000 out已发布 API 费率
Gemini 3.8 / 3.7 / 3.6 Flash375 in · 1,875 out已发布 API 费率
Gemini 3.5 Flash375 in · 2,250 out已发布 API 费率
Gemini 3.5 Flash Lite75 in · 625 out已发布 API 费率
Gemini 2.5 Pro313 in · 2,500 out已发布 API 费率
Gemini 2.5 Flash75 in · 625 out已发布 API 费率
Gemini 2.5 Flash Lite25 in · 100 out已发布 API 费率
Kimi K3750 in · 3,750 out已发布 API 费率
Kimi K2.6 / K2.7 Code238 in · 1,000 out已发布 API 费率
NVIDIA models25–140 in · 70–420 out已发布 API 费率
Questions
How is a token call charged?
Input, cached input, output, and reasoning tokens are measured by the provider. A successful request is charged from actual usage; failed model calls are refunded.
Why can the final cost differ from the estimate?
The estimate reserves enough credit to start safely. The final amount uses the provider's measured usage and returns any unused reservation.