Google released Gemini 3.7 Flash on August 13, positioning it as a model for coding, agents, and work with complex documents. The launch page specified an end date for the introductory API price.
Through December 31, Google listed $0.75 per million input tokens and $3.75 per million output tokens. Starting January 1, 2027, the published rates were $1.50 and $7.50 respectively.
| API usage | Introductory rate per million tokens | Announced January 2027 rate |
|---|---|---|
| Input | $0.75 | $1.50 |
| Output | $3.75 | $7.50 |
These were the rates in the August announcement, not a guarantee that an entire agent task would cost a particular amount. A task may require multiple calls, additional tools, and retries.
Google said 3.7 Flash improved on 3.6 Flash in software engineering and document workflows. Those were vendor performance claims. The rates alone cannot establish the cost of a successful task: retries and corrections affect the amount of usage required.
For a simple budget example, one million input tokens plus one million output tokens would total $4.50 at the introductory rates and $9 at the announced January rates. This is arithmetic from the price table and excludes separate tool charges or other usage.
Google's API pricing documentation lists Search grounding separately, with a shared monthly allowance for Gemini 3 and newer models and charges after that allowance. It also separates inference tiers and caching. A team using search or retained context should therefore budget those line items in addition to the basic input/output calculation.
For unchanged traffic, the announced January 1 rates double the basic token component of the bill.
The model was offered through Google's developer and enterprise tools, with consumer access described through Gemini Spark for eligible Pro and Ultra subscribers. An API price table should not be mistaken for the price of a personal Gemini subscription.
