Google ships Gemini 3.6 Flash and 3.5 Lite

Google has rolled out Gemini 3.6 Flash and 3.5 Flash-Lite across its main AI products.

Google is pushing two new Gemini models into multiple products at once, while the higher-end Pro tier is missing from this release. The rollout already reaches Google AI Studio, Android Studio, Gemini App, Gemini Enterprise, and parts of Google Search.

Google is widening access before it widens ambition

The clearest signal in this release is distribution. Google is not waiting for a single flagship model to anchor the whole Gemini lineup. Instead, it is spreading smaller, faster models across products that already have huge reach.

That matters because distribution changes adoption faster than model naming does. A model that ships inside Android Studio gets used by app developers. A model that lands in Gemini App gets tested by everyday users. A model that appears in Search gets pressure-tested at internet scale.

For developers, the practical path is simple: build with the models that are already in the tools you use every day. For Google, that means more usage data, more feedback, and a cleaner way to separate experimental releases from premium offerings.

• Google AI Studio and Android Studio are the main developer entry points.
• Gemini App gives consumer access without extra setup.
• Gemini Enterprise handles business deployment.
• Google Search is the biggest distribution channel of all.

The missing Pro tier is the real story

The absence of Gemini Pro is more interesting than the launch itself. Google has often used higher-end models to signal capability, but this release makes a different point: the company is prioritizing volume, latency, and product fit over a single prestige model.

That choice suggests a split strategy. Flash and Flash-Lite are for speed and scale. Pro, when it arrives, will likely target heavier reasoning, more demanding enterprise jobs, and users who care more about output quality than response time.

There is also a pricing and infrastructure angle here. Smaller models are easier to place inside consumer apps and search products because they cost less to run per request. That makes them the safer bet when a company wants broad exposure without turning every query into an expensive one.

“We are building AI that is useful for everyone, everywhere,” Sundar Pichai said at Google I/O 2024.

That line from Google CEO Sundar Pichai fits this rollout well. The company is clearly trying to make Gemini feel less like a single model and more like a layer that sits across products.

What developers should compare right now

If you are building on Gemini, the useful comparison is not between marketing names. It is between where each model fits in the stack. Flash models are usually the ones you reach for when speed and cost matter. Pro models matter when you need deeper reasoning and can tolerate slower responses.

That also changes how teams should test. A developer building an Android assistant may care about token latency and battery impact. A search-integrated feature may care about throughput and response consistency. An enterprise workflow may care about policy controls and predictable behavior across large document sets.

• Use AI Studio to prototype model behavior quickly.
• Use Android Studio to test on-device and mobile-first flows.
• Use Gemini Enterprise for managed business usage.
• Use Gemini App to see how the model behaves for normal users.

One practical takeaway: if your product does not need the heaviest model available, you should probably start with the model Google is already shipping everywhere. That gives you faster iteration and a better sense of what users actually notice.

What this rollout means next

Google is making Gemini more like a product family than a single model release. That is a sensible move because different surfaces need different tradeoffs, and the company has enough distribution to support that split.

The open question is whether Pro will arrive as a separate premium tier or as a quieter upgrade inside existing products. If Google keeps pushing Flash and Flash-Lite into more surfaces first, then Pro may become the specialist model rather than the headline act.

For developers, the best move is to test the models that are already live, watch how they behave in real products, and keep an eye on whether Google uses Pro to fill the gap between speed and depth. The next release will tell us whether Gemini is becoming a family with clear roles or a pile of overlapping names.