Gemini 3.7 Flash API Pricing: Per-Token Rates and How to Budget

Gemini 3.7 Flash API pricing is billed per token, with a cheap intro rate now and a higher rate in 2027. Google released the model on August 13, 2026 (Google).

Through December 31, 2026, the API rate is $0.75 per 1M input tokens and $3.75 per 1M output tokens (Google). From January 1, 2027, it becomes $1.50 and $7.50.

This guide shows how to model monthly API cost from your token volume, where to call the model, and how to plan around the intro window. Verify all rates on Google's pricing page.

Gemini 3.7 Flash API Rates Per Token

The Gemini 3.7 Flash API costs $0.75 per 1M input tokens and $3.75 per 1M output tokens during the intro window (Google). That window runs through December 31, 2026.

From January 1, 2027, the regular API rate is $1.50 per 1M input tokens and $7.50 per 1M output tokens (Google). The intro rate is half of the regular rate.

These are Google's published rates and can change without notice. Confirm current numbers on Google's pricing page before you set a budget.

How to Model Monthly API Cost From Token Volume

Model your monthly cost by multiplying input and output tokens by their per-token rate, then adding them. Output is five times the price of input at both rates (Google).

First, estimate tokens per request. Count both the prompt you send and the response you expect back.

Then multiply by your monthly request count. The table below shows three example workloads at the intro rate and the 2027 regular rate.

In our own build work at Layer3 Labs, output tokens are what surprise teams on the first bill. Chatty system prompts inflate input, but long generated answers and code are what move the total.

Where to Call Gemini 3.7 Flash

Developers call Gemini 3.7 Flash through the Gemini API, available in Google AI Studio and Android Studio (Google). Google also offers it via the Gemini Enterprise Agent Platform and the Gemini Enterprise app (Google).

Google AI Studio is the fastest place to test prompts and read the current pricing before you wire up billing. Android Studio brings the model into mobile app development.

Start in AI Studio to confirm the model name and current rate. Then move to production once your token estimates hold up.

Pick the surface that matches your build. AI Studio suits quick prototypes and prompt tuning. The Enterprise Agent Platform suits larger, governed deployments. Confirm which surfaces support 3.7 Flash with Google, since availability can shift.

Three Ways to Cut Your Gemini 3.7 Flash API Bill

Cut your bill by shrinking output tokens first, because output costs five times as much as input at both rates (Google). Ask for shorter answers and cap the response length in your request.

Second, trim your prompts. Long system prompts and pasted context inflate input tokens on every single call.

Third, batch work into the intro window where you can. A one-time evaluation or backfill run costs half as much before December 31, 2026 (Google). Confirm current limits and any caching options with Google before you rely on them.

Budgeting Around the Intro Window

The intro rate ends December 31, 2026, and your API cost doubles on January 1, 2027 (Google). Plan for that step change now.

Front-load heavy batch jobs into the intro window if your work allows it. A backfill, an evaluation run, or a data-labeling pass costs half as much before the deadline (Google).

For ongoing production traffic, budget at the regular rate from day one. That way the January 2027 change does not break your forecast.

A Note on Gemini 3 Pro Pricing

Gemini 3 Pro sits on a higher price tier than 3.7 Flash and is in preview (Google). Its rate is $2 per 1M input and $12 per 1M output, with cached input at $0.20 per 1M (Google).

So Gemini 3 Pro output costs about 1.6 times the 3.7 Flash regular output rate, and more than three times its intro rate (Google). Flash is the cheaper workhorse for high-volume calls.

Reach for Pro when a task needs its extra depth and you accept the higher bill. Use Flash when cost per call matters most. We do not list GPT or Claude API rates here; check each vendor's own pricing page for those.

How to Verify Gemini 3.7 Flash API Pricing

Verify Gemini 3.7 Flash API pricing on Google's official pricing page before you budget or ship. Prices change without notice, and the intro rate ends December 31, 2026 (Google).

Read the rate straight from Google AI for Developers, not from an aggregator. Aggregators can lag behind Google's own changes.

Recheck the page near the end of December 2026. That is when the intro rate expires and the regular rate begins (Google).

Wire your billing forecast to the rate you read on Google's page, not to a cached number in your code. A stale rate in a cost model is how teams get surprised at renewal.