Srpske NovineSrpske Novine

Gemini 3.5 Pro Still in Private Preview, Ten Days to Deadline

Nine days before the end of June, Google's flagship model has yet to reach the Gemini app or AI Studio, despite Pichai's Google I/O promise that general availability arrives in June.

Luka Savić3 min read0
Podeli
Google Gemini model logo on a light background
Gemini 3.5 Pro was unveiled at Google I/O, but broad availability is still lagging./ Gemini / Wikimedia Commons (public domain)
In brief
  • Gemini 3.5 Pro was unveiled at Google I/O on May 19, 2026, with general availability targeted for June.
  • As of June 21, the model remains a limited preview for select Vertex AI enterprise users and has not reached the Gemini app or AI Studio.
  • Prediction markets put the odds of a release by June 30 at roughly 50 to 55 percent.
  • A confirmed context window of 2 million tokens and a Deep Think reasoning mode are part of the package.

Will you actually be able to open the Gemini app and try the new model before the end of June? Right now, the honest answer is probably not. Nine days out from the end of the month, Google’s flagship model Gemini 3.5 Pro is still off-limits to the general public, even though Sundar Pichai promised from the Google I/O stage that general availability would arrive in June.

For context, the model was unveiled on May 19, 2026. Pichai’s words on stage were simple: give us one more month to ship it to you. The month is nearly up, and the deadline is hanging by a thread.

Where exactly is Gemini 3.5 Pro today?

As of June 21, here is the state of play. The model exists, it runs, and it is being tested, but only behind closed doors. It is available in a limited preview to select enterprise users through the Vertex AI platform. That is Google’s service for companies, not something you open on your phone over morning coffee.

In practice, that means ordinary users have no access. The Gemini app most of us carry on our phones is still running an older version. AI Studio, the tool developers use to try out models, has not received the new version either. So everything Pichai showed on stage stays a promise until you can run it yourself.

What do we know about the model itself? Three things are confirmed and look serious. The first is a context window of two million tokens, which roughly means the model can process an enormous amount of text at once, say a whole book or a long thread, without losing the plot. The second is a reasoning mode called Deep Think, built for harder tasks where the model takes more time to think before answering. The third is advanced multimodal ability, meaning it works with text, images, and other formats in the same conversation.

Pricing is its own story. It is expected to follow the earlier ratio, roughly ten times the cost of the lighter Flash version. For companies that makes sense; for the average user it means the strongest version will likely be used sparingly.

Will general availability land on time?

Here the numbers say more than the announcements. Prediction markets, where people bet their own money on the outcome, give a release by June 30 odds of around 50 to 55 percent. That is essentially a coin flip. When a deadline is measured that way, it is clear that even those who follow Google closely are unsure.

The reasons for delay are usually the same. Large models do not break in a demo, they break at scale, when millions of people use them at once. There is a difference between a model that serves a few select firms and one that survives a global audience. It is possible Google simply does not want a bad first impression.

What does that mean for you? If you are waiting to try Gemini 3.5 Pro, sit tight and do not cancel what you already use based on a promise. If you need a strong model these days, lean on what is genuinely available today, not on what is announced for tomorrow.

That leaves one question we will have an answer to in about ten days. Will Google honor its own deadline, or will we be writing another snapshot of the same delay in July?

Podeli
Komentari…
Next storyGoogle Gemini Spark: the AI agent that keeps working after you close the laptopGoogle at its I/O conference on May 19 introduced Spark, an AI assistant that runs in the cloud and performs tasks in Gmail, Docs and Slides without interruption.

From other sections

Most read today

  1. Greece vs Serbia: Zivkovic and Dragojevic Out in Volos
  2. Air Serbia Bars Montenegrin Fans From Larnaca Flight
  3. Student List to Elect Assembly Speaker by Majority Vote
  4. Stevandic re-elected Republika Srpska parliament speaker
  5. Guard says it targeted US radar in Bahrain

Komentari (0)

Ostavi komentar

Komentari prolaze kratku moderaciju. Lični napadi i govor mržnje se brišu.

Još nema komentara. Budi prvi i podeli mišljenje.