benchrInput / 1M$0.75through 2026
Output / 1M$3.75through 2026
Cached / 1M$0.075$0.50 storage / MTok-hour
Context1.05M65,536 max output
Official price table
| Tier | Input / 1M | Output / 1M | Cached input / 1M |
|---|---|---|---|
| Standard · through 2026 | $0.75 | $3.75 | $0.075 |
| Batch / Flex · through 2026 | $0.375 | $1.875 | $0.0375 |
| Priority · through 2026 | $1.35 | $6.75 | $0.135 |
| Standard · from Jan 1, 2027 | $1.50 | $7.50 | $0.15 |
The rate has an expiration date
Google explicitly dates the introductory rates through December 31, 2026. A monthly forecast that crosses January must split token volume between the two periods. Search and Maps grounding share 5,000 free requests or prompts per month, then cost $14 per 1,000; those tool charges are not token rates.
Frequently asked questions
Is there a free API tier?
Yes. Google lists a free tier for Gemini 3.7 Flash, subject to its published limits.
Are Batch and Flex the same price?
During the introductory period both list $0.375 input, $1.875 output, and $0.0375 cached input per million tokens.
Does the output price include thinking tokens?
Google's output pricing covers output tokens, including thinking tokens where used.