API 客戶的優先處理服務
Fast mode offers reliable, high-speed performance with the flexibility to pay-as-you-go. For our latest frontier model, gpt-5.6-sol, you can access up to 2.5x faster speeds with Fast mode.
By choosing Fast mode, you can unlock:
- Predictably low latency: Fast mode generates tokens faster and at a more consistent speed than the Standard processing service, even during peak demand.
- Easy-to-use flexibility: Like Standard processing, Fast mode can be accessed on a flexible, pay-as-you-go basis instead of requiring advance provisioning.
Note: Priority processing was renamed Fast mode on July 30, 2026. You can use either service_tier: priority or service_tier: fast in your API requests.
| 每一百萬個輸入權杖的費用 | 每一百萬個輸入權杖的費用 (快取) | 每一百萬個輸出權杖的費用 | 正常運作時間 SLA3 | 延遲 SLA3 | |
|---|---|---|---|---|---|
GPT-5.6 Sol 不支援長段文字記憶力解功能1 | US$10.00 | US$1.00 | US$60.00 | 99.9% | 每秒 99% > 80 個權杖2 |
GPT-5.6 Terra 不支援長段文字記憶力解功能1 | US$4.00 | US$0.40 | US$24.00 | 99.9% | 每秒 99% > 70 個權杖2 |
GPT-5.6 Luna 不支援長段文字記憶力解功能1 | US$0.40 | US$0.04 | US$2.40 | 99.9% | 每秒 99% > 100 個權杖2 |
GPT-5.5 不支援長段文字記憶力解功能1 | US$12.50 | US$1.250 | US$75.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5.4 mini 不支援長段文字記憶力解功能1 | US$1.50 | US$0.150 | US$9.00 | 99.9% | 每秒 99% > 100 個權杖2 |
GPT-5.4 不支援長段文字記憶力解功能1 | US$5.00 | US$0.500 | US$30.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5.2 | US$3.50 | US$0.350 | US$28.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5.1 | US$2.50 | US$0.250 | US$20.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5 | US$2.50 | US$0.250 | US$20.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5 mini | US$0.45 | US$0.045 | US$3.60 | 99.9% | 每秒 99% > 80 個權杖2 |
GPT-5.1 codex | US$2.50 | US$0.250 | US$20.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-5 codex | US$2.50 | US$0.250 | US$20.00 | 99.9% | 每秒 99% > 50 個權杖2 |
GPT-4.1 | US$3.50 | US$0.875 | US$14.00 | 99.9% | 每秒 99% > 80 個權杖2 |
GPT-4.1 mini | US$0.70 | US$0.175 | US$2.80 | 99.9% | 每秒 99% > 90 個權杖2 |
GPT-4.1 nano | US$0.20 | US$0.050 | US$0.80 | 99.9% | 每秒 99% > 100 個權杖2 |
GPT-4o gpt-4o-2024-11-20 gpt-4o-2024-08-06 | US$4.25 | US$2.125 | US$17.00 | 99.9% | 每秒 99% > 80 個權杖2 |
gpt-4o-2024-05-13 | US$8.75 | — | US$26.25 | 99.9% | 每秒 99% > 80 個權杖2 |
GPT-4o mini | US$0.25 | US$0.125 | US$1.00 | 99.9% | 每秒 99% > 90 個權杖2 |
o3 | US$3.50 | US$0.875 | US$14.00 | 99.9% | 每秒 99% > 80 個權杖2 |
o4-mini | US$2.00 | US$0.500 | US$8.00 | 99.9% | 每秒 99% > 90 個權杖2 |
運作方式
Customers can direct traffic to Fast mode on a per request basis using the existing service_tier parameter, with the option service_tier = "fast".
Tokens served by Fast mode will be billed on a per-token basis, priced at a premium relative to Standard processing rates.
In addition to being configured at the request level, you can also default a project to Fast mode in Project settings > Default Service Tier: Fast. You can still override per request. Selecting Fast in your project settings is equivalent to selecting Priority.
限制說明
- Fast mode rate limits are shared with other service tiers.
- In rare cases, rapid increases to your Fast mode Tokens per Minute can lead to hitting ramp rate limits. If you exceed the ramp rate limit, then additional traffic may be sent to Standard processing instead.