Model Watch · Google
Gemini 3
10 models carry the Gemini 3 name, 9 of them callable today. Here's what separates them, and what each one costs.
From Google's own docs · updated 2026-09-05 10:32 UTC
The family started with Gemini 3 Flash Preview on 2025-12-17 and the newest is Gemini 3.8 Flash from 2026-09-02. Running one of these for a month costs anything from $0.74 on Gemini 3.1 Flash-Lite to $5.92 on Gemini 3.1 Pro Preview, for identical work.
Every Gemini 3 model
| Model | Status | Released | Context | In / 1M | Out / 1M | Your month |
|---|---|---|---|---|---|---|
| Gemini 3.8 Flash | released | 2026-09-02 | 1M | $0.75 | $3.75 | $1.92 |
| Gemini 3.7 Flash | released | 2026-08-13 | 1M | $0.75 | $3.75 | $1.92 |
| Gemini 3.6 Flash | released | 2026-07-21 | 1M | $0.75 | $3.75 | $1.92 |
| Gemini 3.5 Flash-Lite | released | 2026-07-21 | 1M | $0.30 | $2.50 | $1.17 |
| Gemini 3.5 Flash | released | 2026-05-19 | 1M | $1.50 | $9.00 | $4.44 |
| Gemini 3.1 Flash-Lite | released | 2026-05-07 | 1M | $0.25 | $1.50 | $0.74 |
| Gemini 3.1 Pro Preview | preview | 2026-02-19 | 1M | $2.00 | $12.00 | $5.92 |
| Gemini 3 Flash Preview | preview | 2025-12-17 | 1M | $0.50 | $3.00 | $1.48 |
| Gemini 3.5 Pro | announced | — | — | — | — | — |
| Gemini 3.1 Deep Think | released | — | — | — | — | — |
The highlighted row is the newest one Google has actually released. The monthly figure is 700 words in and 500 out, twenty times a day, at list prices. Change the workload.
Which one should you be using?
Gemini 3.8 Flash is the newest released model in the family, out on 2026-09-02. Newer is not automatically better for your job, but it is usually the one the lab is optimising, and on this table it is rarely the most expensive.
Why there are so many numbers
A point release is a new model, not a patch. Google ships a new one under the same family name when the training or the tuning changes, and the old one usually keeps working at its old price until it's deprecated.
That's why two models in this table can hold the same amount of text and cost different amounts. You're picking a specific model, not a family, and the API name in each model's page is what actually goes in your code.
What each one is for
- Gemini 3.8 FlashGoogle's newest stable model, and it's a Flash. The price is introductory: it doubles on 1 January 2027, so budget for $1.50 in and $7.50 out.
- Gemini 3.7 FlashThree weeks older than 3.8 Flash at exactly the same price. Same introductory rate, same January cliff.
- Gemini 3.6 FlashJuly's Flash. Google shipped three of these in seven weeks, all priced identically.
- Gemini 3.5 Flash-LiteThirty cents a million on a million-token window. The cheapest Gemini you can call that isn't a generation behind.
- Gemini 3.5 FlashOlder and dearer than every Flash that followed it, because it never got the introductory rate. Move off it.
- Gemini 3.1 Flash-LiteCheaper than the 3.5 Lite that replaced it, and audio costs double the text rate. Worth knowing if your volume is text.
- Gemini 3.1 Pro PreviewThe only Pro-tier Gemini you can call, and Google still labels it preview seven months in. Prompts over 200K tokens bill at double.
- Gemini 3 Flash PreviewThe first Gemini 3 Flash, still sitting in preview while four stable ones shipped past it.
- Gemini 3.5 ProGoogle says "3.5 Pro coming soon" and nothing else. No id, no price, no date, and the Flash line has already run four versions past it.
- Gemini 3.1 Deep ThinkIt exists, it's in the Gemini app, and that's all Google will tell you. No API id and no price anywhere on their docs.
Keep reading
Everything Google has published · Google prices in full · find a model by budget · compare any two