Batch API 价格与节省

Batch API 以异步方式运行请求（结果通常在 24 小时内返回），换取更低价格。适合：评测、分类、归档摘要、批量生成 Embedding。不适合：用户正在等待结果的任务。

模型	常规输入 /1M	Batch 输入 /1M	节省
GPT-6.1 Sol	$2.00	$1.00	50%
Claude Sonnet 5.5	$2.00	$1.00	50%
Claude Opus 5.5	$4.00	$2.00	50%
GPT-6 Sol	$2.00	$1.00	50%
GPT-6 Astra	$10.00	$5.00	50%
Claude Fable 5.1	$10.00	$5.00	50%
Claude Opus 5	$5.00	$2.50	50%
GPT-5.6 Sol	$4.00	$2.00	50%
GPT-5.6 Terra	$2.00	$1.00	50%
Claude Sonnet 5	$2.00	$1.00	50%
Claude Mythos 5	$10.00	$5.00	50%
Claude Fable 5	$10.00	$5.00	50%
Claude Opus 4.8	$5.00	$2.50	50%
Gemini 3.5 Flash	$1.50	$0.75	50%
GPT-5.5	$5.00	$2.50	50%
Claude Opus 4.7	$5.00	$2.50	50%
GPT-5.4	$2.50	$1.25	50%
Gemini 3.1 Pro Preview	$2.00	$1.00	50%
Claude Sonnet 4.6	$3.00	$1.50	50%
Claude Opus 4.6	$5.00	$2.50	50%
GPT-5.2	$1.75	$0.875	50%
Claude Opus 4.5	$5.00	$2.50	50%
GPT-5.1	$1.25	$0.625	50%
Claude Sonnet 4.5	$3.00	$1.50	50%
GPT-5	$1.25	$0.625	50%
Gemini 2.5 Pro	$1.25	$0.625	50%
o3	$2.00	$1.00	50%
GPT-4.1	$2.00	$1.00	50%
GPT-4o	$2.50	$1.25	50%
Claude Mythos 5.1	$10.00	$5.00	50%
Gemini Pro (latest)	$2.00	$1.00	50%
Claude Haiku 5.5	$0.10	$0.05	50%
GPT-6 Luna	$0.10	$0.05	50%
Gemini 3.8 Flash	$0.75	$0.375	50%
Gemini 3.7 Flash	$0.75	$0.375	50%
Gemini Flash (latest)	$0.75	$0.375	50%
Gemini 3.5 Flash Lite	$0.30	$0.15	50%
Gemini 3.6 Flash	$0.75	$0.375	50%
Gemini Flash Lite (latest)	$0.30	$0.15	50%
GPT-5.6 Luna	$0.20	$0.10	50%
Gemini 3.1 Flash Lite	$0.25	$0.125	50%
GPT-5.4 mini	$0.75	$0.375	50%
GPT-5.4 nano	$0.20	$0.10	50%
Gemini 3 Flash Preview	$0.50	$0.25	50%
Claude Haiku 4.5	$1.00	$0.50	50%
GPT-5 mini	$0.25	$0.125	50%
GPT-5 nano	$0.05	$0.025	50%
Gemini 2.5 Flash	$0.30	$0.15	50%
Gemini 2.5 Flash Lite	$0.10	$0.05	50%
GPT-4.1 mini	$0.40	$0.20	50%
GPT-4o mini	$0.15	$0.075	50%
Grok 4.3	$1.25	$1.00	20%
Grok 4.20 Non Reasoning	$1.25	$1.00	20%
Grok 4.20 Reasoning	$1.25	$1.00	20%

常见问题

Batch 输出也打折吗？

通常是的，折扣比例与输入相同。每个模型页面都列出了 Batch 输入和输出价格。

Batch 能和提示缓存一起用吗？

部分提供商允许，并对 Batch 中的缓存 Token 按优惠价计费；具体规则请查看提供商文档。

哪些提供商这里没有 Batch 价格？

我们数据中未公布 Batch 价格的模型不在此表中。