发布于 2026-08-16 · 赵予安
直接回答
Google Gemini 3.5 Flash 标价为每百万 token 0.217 / 1.302，上下文 1,048,576。第一方缓存分数未公布。 这篇指南面向正在比较模型质量、成本、路由策略和生产上线风险的产品与平台团队。
What shipped
Google Gemini 3.5 Flash has appeared in the NextModel catalog as `google/gemini-3.5-flash`. The display name is "Google: Gemini 3.5 Flash" and the provider is Google. The catalog lists the public slug as `google--gemini-3.5-flash`. The catalog snapshot is dated 2026-08-16.
Pricing and context
The listed input price is $0.217 per 1M tokens. The listed output price is $1.302 per 1M tokens. The context window is 1,048,576 tokens. The latency band is 1000 to 3000 ms. Related catalog entries include Gemini 3 Flash Preview at $0.072 input and $0.434 output, Gemini 2.5 Flash at $0.043 input and $0.362 output, and Gemini 3.5 Flash Lite at $0.043 input and $0.362 output.
Who this launch is for
This model has streaming, tool, JSON, vision, and long context capabilities. The catalog lists it for vision, long-context, and agent use cases.
Still unverified
The release date is unpublished in this catalog snapshot. Independent third-party quality scores are not in these notes. No first-party CacheSafety Bench row exists for this exact model in the NextModel bench table. Safe Hit, Bad Hit, and trap rates are not in these notes.