Gemini 3.8 Flash Pricing Is Locked Through 2026 — Here’s What Doubles on January 1, 2027

By Alex Kim · Tech & Digital Editor · Last updated September 21, 2026 · 3 sources checked

Where to Look First

Gemini 3.8 Flash launched on September 2, 2026, and the model itself isn’t the story here. The pricing structure attached to it is. Google tied the launch to an introductory rate that quietly has an expiration date, and that date lands square in the middle of most teams’ Q1 2027 budget planning.

Model launches usually come with a benchmark chart and a shrug about pricing. This one is different because the pricing itself is the news — a fixed-length discount window with a hard cutoff, attached to three different model versions at once, doubling on a single calendar day rather than fading out gradually.

What Changed With Gemini 3.8 Flash Pricing on September 2, 2026?

Google set Gemini 3.8 Flash at $0.75 per million input tokens and $3.75 per million output tokens starting on its September 2, 2026 launch date, according to Google’s official Gemini API pricing page. That’s the same intro rate Google applied to Gemini 3.6 Flash and 3.7 Flash before it.

Here’s the part that’s easy to miss on a quick skim: this isn’t the permanent rate. It’s a window. When I checked the official pricing documentation directly, the fine print attaches an explicit end date — December 31, 2026 — to the $0.75/$3.75 numbers, and a new rate takes over the next day.

Short version: you’re not looking at ‘Gemini 3.8 Flash pricing.’ You’re looking at launch-window pricing with four months left on the clock.

How Much Will Gemini 3.8 Flash Cost After the Price Doubles on January 1, 2027?

Starting January 1, 2027, Gemini 3.8 Flash moves to $1.50 per million input tokens and $7.50 per million output tokens — a straight doubling on both sides of the meter, confirmed on Google’s pricing page and independently reported by Enterprise DNA’s AI Pulse coverage on September 3, 2026.

It’s not just the headline rate, either. The batch, flex, and priority multipliers are set as percentages of the base price, so they double right along with it. There’s no separate grace period tucked into those tiers.

For a team running Gemini 3.8 Flash at meaningful volume, that’s not a rounding error. A workload that costs $3,000 a month in December 2026 costs $6,000 in January 2027 with zero change in usage.

Gemini 3.8 Flash Pricing at a Glance

Laid out side by side, the before-and-after is blunt:

Is Gemini 3.8 Flash Still Cheaper Than Gemini 3.7 Flash Right Now?

Right now, no — they’re priced identically. Google kept 3.6 Flash, 3.7 Flash, and 3.8 Flash on the same $0.75/$3.75 intro rate, so switching to 3.8 Flash today costs the same per token as staying on 3.7 Flash.

That changes the calculus most teams run when a new model drops. Usually the question is ‘is the new model worth the price premium?’ Here there isn’t one. When I compared the two rate cards line by line, the only real trade-off left is capability and latency, not cost — our earlier look at Gemini 3.8 Flash’s response times found it noticeably slower than 3.7 Flash for some workloads, which matters more when price isn’t the deciding factor.

So the pricing question isn’t ‘which model is cheaper.’ It’s ‘how long does this shared rate last,’ and the answer is the same 116 days for both.

How Do You Actually Lock In Current Pricing Before the Increase Hits?

You can’t freeze the rate itself — there’s no contract or commitment tier that holds $0.75/$3.75 past January 1, 2027. What you can control is when tokens get processed.

• Front-load high-volume, non-urgent jobs (batch summarization, embeddings refreshes, backfills) into Q4 2026 while the lower rate applies.
• Re-run your cost model with the January 2027 numbers now, not in December, so budget approval isn’t a scramble.
• If you’re mid-evaluation between 3.7 Flash and 3.8 Flash, decide before the deadline — after January 1 the pricing is identical again, so you lose nothing by testing both through December.
• Check whether your workload actually needs Flash-tier latency, or whether a cheaper, slower batch tier absorbs some of the increase.

This part trips a lot of teams up.

People assume ‘intro pricing’ means a small discount. A full doubling is a different category of change, and it deserves a calendar reminder, not a footnote.

What Does This Actually Cost a Small App at Different Usage Levels?

Take a mid-size chatbot processing 50 million input tokens and 10 million output tokens a month — a realistic load for a customer support tool with a few thousand daily users.

At the small end, an extra $75 a month barely registers. At the 2 billion token end, it’s a line item finance will ask about. I found that the gap scales linearly with no volume discount kicking in to soften it — the doubling applies the same whether you’re a solo developer or processing at enterprise scale.

Common Mix-Ups With the Intro Pricing Window

A few things people get wrong when they first read about this change:

Thinking it’s Gemini 3.8 Flash-specific. It isn’t. The same intro window and the same doubling apply to 3.6 Flash and 3.7 Flash, so switching between them doesn’t dodge the increase.

Assuming enterprise contracts are exempt. Nothing in the public pricing documentation carves out an exception for committed-use or enterprise agreements on this particular rate card.

Treating the January date as tentative. Google’s own pricing page states the deadline as a fixed calendar date, not a placeholder — there’s no language suggesting it could slide, so the safer planning move is building budgets around January 1, 2027 now rather than waiting to see if it shifts.

Honestly, the most common mistake is just not noticing the date at all. Most pricing pages get bookmarked once and never revisited until the invoice jumps.

Frequently Asked Questions

Does the price increase apply to Gemini 3.7 Flash and 3.6 Flash too?

Yes. Google’s intro pricing of $0.75 and $3.75 per million tokens covers Gemini 3.6 Flash, 3.7 Flash, and 3.8 Flash together, and all three move to $1.50 and $7.50 per million tokens on January 1, 2027.

Do batch and priority requests get the same discount before the deadline?

Yes, at their usual discounted percentage of the base rate — but that base rate is what doubles on January 1, 2027, so batch and priority pricing doubles along with it. There’s no separate grace period for those tiers.

Is there any way to keep the September 2026 rate after the deadline?

No public mechanism lets you freeze the old rate past January 1, 2027. The only lever available now is timing: tokens processed before the deadline bill at the lower rate, and nothing after it does.

Sources

• Google AI for Developers — Gemini API pricing
• Google Cloud — Gemini Enterprise generative AI pricing
• Enterprise DNA AI Pulse — Google locks Gemini 3.8 Flash pricing through year-end

Fact-checked based on public sources as of September 21, 2026. Pricing is subject to change — always confirm current rates on Google’s official pricing page before budgeting.