Gemini 3.8 and 3.7 Flash introductory pricing
Gemini 3.8 Flash is $0.75 input / $3.75 output per million tokens through 31 December 2026 on the Developer API. After that date Google lists a planned step-up to $1.50 / $7.50. This calculator uses the current intro rate and flags the sunset in the pricing-table notes.
Gemini 3.7 Flash (`gemini-3.7-flash`) is on the same intro tier: $0.75 / $3.75 through 31 December 2026, then $1.50 / $7.50. That figure is Google Cloud Agent Platform global pricing; the Developer API card fetched on 8 September 2026 did not list 3.7. Non-global Agent Platform regions are 10% higher.
Gemini 3.1 Pro is $2 / $12 per million for prompts of 200,000 tokens or fewer. Gemini 2.5 Pro remains $1.25 / $10 for the same short-context band; prompts above 200k on 2.5 Pro step up to $2.50 / $15. This calculator uses the ≤200k tier — if you routinely stuff more than 200k tokens into a single prompt, our Pro rows will understate that case.
Gemini 2.5 Flash is $0.30 per million for text, image, or video input and $2.50 per million output. Audio input is billed at a higher $1.00 / 1M and is not selected automatically here. Gemini 2.5 Flash-Lite ($0.10 / $0.40) is omitted from this packed table.
What v1 does not price
Context caching, Batch (~50% off), Flex, and Priority tiers are separate columns on the official pricing page. Grounding with Google Search or Maps is priced per grounded prompt after a free daily allowance — it will never appear in a pure token formula. Computer-use and native-audio live APIs have their own meters.
Google’s tables state that output price includes thinking tokens on 2.5 Pro. If you enable thinking budgets, the output field in this form should reflect that larger number. A “short answer” mental model will undershoot when reasoning is on.
Image generation, TTS, and live translation endpoints are not text-in/text-out at these rates. If your app is multimodal on the output side, this calculator is the wrong tool until those rows are added to the config. Vertex AI SKUs and committed use are a different card than the Developer API rates shown here.
How to use the table for a Gemini budget
Start with a log-based average: input tokens and output tokens per successful request, including thinking. Set requests per day to production QPS converted to a day, not to a peak minute. Turn on monthly projection for a 30-day sketch. If Pro is only needed for hard cases, uncheck it and price Flash for the bulk path, then add a second share link for the Pro share of traffic.
Free-tier Gemini is not $0 in production. Limits, data-use terms, and paid-tier rate cards differ. This site always uses paid standard rates so a hobby quota does not leak into a board deck.
When Google ships a price change, edit the Gemini objects in src/data/pricing.ts and set lastUpdated. Recheck 3.8 and 3.7 Flash before you lock a 2027 budget — intro pricing ends 31 December 2026.
Frequently asked questions
How much does Gemini 3.8 Flash cost per million tokens?
Introductory Developer API pricing through 31 December 2026: $0.75 input and $3.75 output per million tokens. Google lists a planned step-up to $1.50 / $7.50 after that date.
How much does Gemini 3.7 Flash cost per million tokens?
Same intro tier as 3.8 Flash on Google Cloud Agent Platform (global): $0.75 input and $3.75 output per million tokens through 31 December 2026, then $1.50 / $7.50.
How much does Gemini 3.1 Pro cost?
$2.00 input and $12.00 output per million tokens on the standard text tier for prompts ≤ 200k. Longer prompts are billed higher and are not shown in this row.
How much does Gemini 2.5 Pro cost per million tokens?
On the standard paid tier in this snapshot: $1.25 input and $10.00 output per million tokens for prompts ≤ 200k. Longer prompts are $2.50 / $15.00. Output includes thinking tokens.
Does this Gemini calculator include Search grounding?
No. Grounding is billed per grounded prompt after a free allowance. Add that on top if your app enables Google Search or Maps grounding.