Gemini 3.5 Pro Release Date: Features, Price, and What’s Confirmed

When is Gemini 3.5 Pro coming out?

Gemini 3.5 Pro is not generally available yet. Google announced it at Google I/O on May 19, 2026, but shipped only Gemini 3.5 Flash that day. Pro went into limited preview, with general availability targeted for sometime in June 2026 and no specific date committed.

On stage, Sundar Pichai told developers to “give us until next month” before Pro would reach them, a line widely reported to have drawn an audible groan from the audience. As of June 23, that is still where things stand: internal use and a limited enterprise preview, with the public release pending.

If you want the full rundown of what else launched that day, see our Google I/O 2026 recap.

Is Gemini 3.5 Pro out yet?

No. The model is available only to a small set of enterprise customers through Google’s Vertex AI preview. It is not in the consumer Gemini app, not in Google AI Studio’s public model picker, and not in the general API as a stable release.

Google’s official Gemini blog framed Pro as “being used internally” with a rollout planned for the following month, and nothing has formally superseded that. Until the model is generally available and independent testers can run their own evaluations, its advertised capabilities remain Google’s claims rather than verified results.

How likely is a June arrival? Prediction markets have placed the odds of a release by June 30 at roughly 50 to 55 percent. In other words, leaning yes, but not a certainty. Treat any post claiming a specific launch day with caution, because Google has not committed to one.

What features and specs are confirmed?

Here it helps to separate what Google has stated from what is still an estimate. The targets below come from Google’s own positioning, but the verified numbers only arrive with the model card at general availability.

Confirmed targets:

• A 2-million-token context window, double Gemini 3.5 Flash’s 1 million and the largest of any production frontier model at announcement.
• A “Deep Think” reasoning mode built for harder, multi-step analysis.
• Frontier multimodal understanding across text, images, and other formats.
• Positioning as the successor to the old Gemini Ultra tier, absorbing the hardest reasoning and long-context work.

The reason Pro matters is the gap Flash left. Flash shipped strong on coding and agentic benchmarks, beating the previous Gemini 3.1 Pro, but it gave ground on the hardest reasoning and long-context recall. Pro is Google’s answer to exactly that gap.

Gemini 3.5 Pro vs Gemini 3.5 Flash

If you need something today, Flash is the live option. Here is how the two tiers compare based on confirmed information.

Flash is the default in the consumer Gemini app and in Search’s AI Mode, so most people are already using 3.5 without paying anything.

How much will Gemini 3.5 Pro cost?

Google has not announced Pro pricing. The confirmed reference point is Flash at $1.50 per million input tokens and $9 per million output tokens, and Pro is expected to sit above that.

Several outlets have estimated consumer access through a roughly $20 per month plan and a higher tier around $250 per month, with the Deep Think mode possibly limited to the top tier. Treat those figures as informed expectations, not Google’s published prices. The real numbers, the API model string, and the context limits all land together in the model card at general availability.

For developers planning ahead, context caching matters more than the headline rate. On long-context workloads where the same document or system prompt repeats, cached input pricing is often the difference between a viable deployment and an unworkable bill.

How does it compare to Claude and GPT-5.5?

Pro is aimed squarely at the frontier tier where Anthropic’s Claude and OpenAI’s GPT models have set the pace. The competitive question is whether a 2-million-token window and a dedicated reasoning mode let Google lead on long-context and hard-reasoning tasks, not just on speed.

Until independent benchmarks exist, that remains open. For a current side-by-side of the major assistants, see our Claude vs ChatGPT vs Gemini comparison, and for the coding angle specifically, our best AI coding assistants guide. You can also see where each model sits in our top AI models overview.

Should you wait for Gemini 3.5 Pro?

For most people and most tasks, there is no reason to wait. For a specific set of heavy workloads, Pro could be worth holding out for.

Reasons to wait for Pro

• You need the full 2-million-token context to load entire codebases, long contracts, or large document sets.
• Your work is reasoning-heavy and accuracy matters more than speed.
• You are an enterprise standardizing on Google’s frontier tier and want the strongest option.

Reasons not to wait

• Your work is coding or agent-leaning, where Flash already beats last year’s Pro at lower cost.
• You mostly do writing, research, and everyday questions, which Flash handles well for free.
• You need certainty now, since Pro’s date, price, and verified performance are all still pending.

If you are weighing Google against Apple’s assistant for everyday use, our Gemini vs Apple Intelligence breakdown is a useful companion.

How to get access when it launches

When Pro ships, it tends to appear quietly rather than with a countdown. A few practical ways to be first.

For developers, watch the Google AI Studio model picker, where new models usually appear the moment the API is ready, and keep an eye on the official Gemini API changelog. For enterprise teams, you can request allowlist access through the Vertex AI Model Garden by searching for the Pro model, or ask your Google Cloud account team. Expect the first release to carry a preview-style model name before the stable identifier follows.

The bottom line

Gemini 3.5 Pro is one of the most anticipated model launches of the year, and as of June 23, 2026 it is still not here. Google has committed to a June window and a strong feature set, led by a 2-million-token context and Deep Think reasoning, but the price and verified performance only become real at general availability. Until then, Gemini 3.5 Flash is the capable, live option, and the smart move is to test Flash now and re-evaluate the day Pro’s model card lands.

FAQ

Is Gemini 3.5 Pro available yet?

Not for the general public. As of June 23, 2026, Gemini 3.5 Pro is in limited preview for select enterprise customers through Google’s Vertex AI platform. The public-facing 3.5 model is Gemini 3.5 Flash, which is live in the Gemini app, Search’s AI Mode, and the API.

When is Gemini 3.5 Pro coming out?

Google announced it at I/O on May 19, 2026 and targeted general availability for June 2026, without committing to a specific date. As of late June, it has not shipped publicly, and prediction markets put the odds of a release by June 30 at roughly 50 to 55 percent.

Is Gemini 3.5 Pro free?

Likely not, beyond limited trials. Gemini 3.5 Flash is free in the Gemini app and Search by default, while Pro is expected to sit behind paid consumer plans and metered API pricing. Google has not published Pro’s pricing, so any specific figures are estimates until general availability.

Does Gemini 3.5 Pro really have a 2-million-token context window?

A 2-million-token context window is Google’s stated target for Pro, double Gemini 3.5 Flash’s 1 million. It would be the largest of any production frontier model at announcement. The confirmed number will be published in the model card when the model becomes generally available.

What is the difference between Gemini 3.5 Pro and Flash?

Flash is the fast, lower-cost model that is live now and serves as the default Gemini experience. Pro is the higher-capability tier built for the hardest reasoning, long-context, and multimodal tasks. Flash handles everyday and high-volume work; Pro targets the most demanding jobs.

Is Deep Think mode only on the most expensive tier?

Reporting suggests Deep Think may be limited to Google’s top subscription tier, but Google has not confirmed this. The official details on which features land on which plans will come with the general availability announcement.