Gemini 3.7 Flash costs $0.75 per million input tokens and $3.75 per million output tokens, with cached input at $0.075, read from the Google AI pricing page on 16 August 2026.
Update, 3 November 2026: Gemini 3.7 Flash has been retired. Google’s deprecations page, read on 3 November 2026, says gemini-3.7-flash is no longer available in Google AI Studio; the endpoint is not shut down but automatically routes API requests to gemini-3.8-flash, and the pricing page no longer carries a line for 3.7 Flash. The same applies to Gemini 3.5 Flash, whose requests now route to gemini-3.6-flash. Gemini 3.8 Flash is listed at the same rates as 3.7 Flash was: $0.75 input, $0.075 cached and $3.75 output through 31 December 2026, then $1.50, $0.15 and $7.50 from 1 January 2027. The figures below are kept as the record of what 3.7 Flash cost; for new work, read our page on Gemini 3.8 Flash pricing.
Those are promotional rates that Google publishes as applying through 31 December 2026, after which the same page lists $1.50 and $7.50: double, on every line. There is also a free tier.
What Gemini 3.7 Flash costs, on both grids
Google publishes two columns for this model, separated by a date rather than by a prompt length or a plan.
| Billing line | Through 31 Dec 2026 | From 1 Jan 2027 |
|---|---|---|
| Input | $0.75 | $1.50 |
| Cached input | $0.075 | $0.15 |
| Output | $3.75 | $7.50 |
Gemini 3.7 Flash at $0.75 in and $3.75 out per million, read 16 August 2026
Gemini 3.6 Flash carries exactly the same two grids and the same date, so switching between the two versions changes nothing about what you pay.
A rate with a date is not a rate
This is the part worth being careful about, and we have been caught by it in both directions.
A scheduled increase is an announcement, not a fact. In this same period, Anthropic had published that Claude Sonnet 5 would rise from $2 and $10 to $3 and $15 on 1 September 2026, then cancelled the increase and made the launch rates standard. Anyone who had migrated away from Sonnet 5 to avoid that rise would have moved for nothing.
The reverse risk is just as real. Budgeting a year of Gemini 3.7 Flash at $0.75 assumes a discount you have been told will end. The sane position is to plan at $1.50 and treat the current rate as money you are saving until the date, rather than as the price of the model.
Budget at the higher figure and bank the difference. A published rate with a published expiry is two numbers, and only one of them is a commitment.
Where it sits in the Flash line
Google runs several Flash and Flash-Lite models at once, and the newest is not the dearest.
| Model | Input / cached / output, per million |
|---|---|
| Gemini 3.8 Flash, promotional (replaces 3.7 Flash) | $0.75 / $0.075 / $3.75 |
| Gemini 3.7 Flash, promotional (retired, routes to 3.8 Flash) | $0.75 / $0.075 / $3.75 |
| Gemini 3.7 Flash, from 2027 | $1.50 / $0.15 / $7.50 |
| Gemini 3.6 Flash, promotional | $0.75 / $0.075 / $3.75 |
| Gemini 3.5 Flash (retired, routes to 3.6 Flash) | $1.50 / $0.15 / $9.00 |
| Gemini 3.5 Flash-Lite | $0.30 / $0.03 / $2.50 |
| Gemini 3.1 Flash-Lite | $0.25 / $0.025 / $1.50 |
Two readings come out of that table. On the promotional rate, 3.7 Flash undercuts 3.5 Flash by half on input and by more than half on output. And even after 1 January 2027 it stays cheaper than 3.5 Flash on output, at $7.50 against $9.00, for the same input price. So the newer model is the better deal on both sides of the date, which is not something you can assume in this market, as the generation that followed it would later demonstrate.
The Flash-Lite models sit well below all of them. If your task is classification, extraction or routing, 3.1 Flash-Lite at $0.25 and $1.50 does that work at a third of the price, and we mapped where the floor sits across every provider.
Audio is billed separately, and it is not close
The rates above are for text, image and video input. Audio input on the Flash line is charged on its own line and it is substantially dearer: on Gemini 3.1 Flash-Lite, audio input costs $0.50 per million against $0.25 for text, exactly double, with the cached rate doubling to match. For those interested in continuous audio interactions, see also Gemini Live audio by the hour.
That distinction disappears from almost every comparison table, ours included until we checked. If your product transcribes or listens, the headline text rate tells you very little about your bill, and our audio API pricing page covers what the specialised transcription services charge for the same work.
What a month actually costs
Three shapes of work on the promotional rate, with no caching, which is the honest default for traffic that does not repeat.
| Monthly workload | Now / from 2027 |
|---|---|
| Assistant: 2M in, 0.5M out | $3.38 / $6.75 |
| Agent: 10M in, 2M out | $15.00 / $30.00 |
| Heavy: 50M in, 10M out | $75.00 / $150.00 |
The doubling is exact at every volume, because both lines scale by the same factor. Whatever you are paying today, write down twice that number and check it still works. Our LLM cost calculator prices the same volume across every provider we track.
The free tier, and what it is for
Gemini 3.7 Flash had a free tier when we read the page in August, as did 3.6 Flash, 3.5 Flash, both Flash-Lite models and Gemini 2.5 Pro; 3.7 Flash and 3.5 Flash have since been retired. The models without one are the previews and the generation models: 3.1 Pro Preview, the image models, Veo and Lyria.
A free tier is a development convenience, not a production plan, and it comes with rate limits and data terms that differ from the paid tier. We treat it as a way to try a model rather than a way to run one, and we go through what free tiers actually give you across the market on our page on the cheap and free end of the API market.
Frequently asked questions
Is Gemini 3.7 Flash still available?
Not as a model you can pick. Since its retirement, Google AI Studio no longer offers it and API calls to gemini-3.7-flash are routed to gemini-3.8-flash, according to the deprecations page read on 3 November 2026. Code that still names 3.7 Flash keeps working, but it is served by 3.8 Flash, listed at the same rates through 31 December 2026.
How much does Gemini 3.7 Flash cost per million tokens?
$0.75 for input and $3.75 for output, with cached input at $0.075, through 31 December 2026. From 1 January 2027 the published rates are $1.50, $7.50 and $0.15. Read from the Google AI pricing page on 16 August 2026.
Will the price really double in January 2027?
That is what Google publishes today. It is an announcement rather than a commitment: Anthropic cancelled a comparable scheduled increase on Claude Sonnet 5 during the same period. Budget at the higher figure and re-check the source before the date.
Is Gemini 3.7 Flash cheaper than Gemini 3.5 Flash?
Yes, on both grids. On the promotional rate it is half the input price and less than half the output. Even from 2027 it matches 3.5 Flash on input and undercuts it on output, at $7.50 against $9.00.
Does Gemini 3.7 Flash have a free tier?
Yes. So do 3.6 Flash, 3.5 Flash and both Flash-Lite models. The previews and the image, video and music generation models do not.
Is audio billed at the same rate as text?
No. Audio input is a separate and higher line across the Flash family: on 3.1 Flash-Lite it is $0.50 per million against $0.25 for text, image and video.
What we verified and what we did not
Verified: input, cached input and output rates for Gemini 3.7 Flash, 3.6 Flash, 3.5 Flash, 3.5 Flash-Lite and 3.1 Flash-Lite, on both the promotional and the post-2026 grids, the separate audio input rates, and which models carry a free tier, all read from the Google AI pricing page on 16 August 2026.
Not verified, and therefore not asserted: the rates on Vertex AI, which Google prices separately from the Gemini API; what the free tier’s rate limits actually allow, which the pricing page does not state; and any benchmark or quality claim, since this page prices models rather than ranking them.
This page carries an expiry date that we did not choose, which makes it the clearest example on this site of why every figure here is stamped with the day it was read. Check the source before you commit a budget past December.