Model launches
Where is Gemini 3.5 Pro? Google's 'next month' promise, explained
Google teased a powerful new Gemini at I/O. Here's why you still can't use it — and what you can.
The answer
Google announced Gemini 3.5 Pro in May 2026 but hasn't released it yet — only Flash is out.
If you saw headlines in May about a powerful new Google AI called Gemini 3.5 Pro and have been waiting to try it, here's a plain-English guide to what's actually going on — and what you can use right now.
What Google actually said
At Google I/O — Google's annual technology event — on 19 May 2026, Google introduced Gemini 3.5. It released Gemini 3.5 Flash, a fast, low-cost model, and made it the default in the Gemini app and in Google Search's AI Mode straight away. At the same time, it announced Gemini 3.5 Pro, the bigger and more capable version — but it didn't release it. Google said Pro was already being used inside the company and would roll out 'next month'. Said in mid-May, that pointed to June.
By late June 2026, 'next month' had essentially arrived and gone, and Gemini 3.5 Pro still wasn't something the public could use. It wasn't in the Gemini app, it wasn't in Google AI Studio (where developers test models), and it wasn't in the main, stable programming interface (the API). In short: announced, talked about, but not yet in your hands.
About those '2 million tokens' you may have read
A lot of articles list impressive specs for Gemini 3.5 Pro — most often a 2-million-token context window (roughly, how much text it can consider at once) and a 'Deep Think' reasoning mode. It's worth knowing that these come from reports and expectations, not from Google's confirmed details for Pro. They might turn out to be right — but until Google publishes an official technical report (a 'model card'), they're best treated as educated guesses. For comparison, the Flash model that is out has a context of about 1 million tokens, according to Google's own materials.
Trade coverage described Gemini 3.5 Pro as nearing a June launch with a roughly 2-million-token context window and a 'Deep Think' reasoning mode — expectations Google had not confirmed.
Why the wait is normal
Here's the reassuring part: a bigger model arriving after a smaller one is completely normal. The larger and more capable a model is, the more testing, safety work and behind-the-scenes engineering it takes to get it ready for millions of people. Releasing the fast, cheap model first — so most users get an upgrade quickly — and holding the heavyweight back to polish it is a sensible plan, not a red flag. The only thing the delay really tells us is that the 'next month' estimate was a bit optimistic, which is common when a date is given from a stage.
Google announced Gemini 3.5 at I/O on 19 May 2026, shipping Flash as the default across the Gemini app and Search while describing the larger Pro model as in internal use and rolling out the following month.
What should you do in the meantime? If you're a everyday user, you're already getting Flash in the Gemini app and in Search — no action needed. If you're a developer or business deciding what to build on, Flash is the available, documented option; treat Pro as 'coming, eventually' rather than something you can schedule around. When Pro actually launches — with a real release, a model card and benchmark numbers — that's the moment to compare it properly. Good tech is worth reviewing when it's actually out, and this part isn't yet.
Frequently asked questions
Can I use Gemini 3.5 Pro right now?
Didn't Google promise it in June?
Is it true that Gemini 3.5 Pro has a 2-million-token memory?
Should I be worried that it's late?
What's the best Gemini I can use today?
Sources
- Gemini 3.5: frontier intelligence with action — Google, 19 May 2026
- 100 things we announced at Google I/O 2026 — Google, 19 May 2026
- Google Gemini 3.5 Pro Nears June Launch With 2 Million Token Context And Deep Think Reasoning — Tech Times, 6 June 2026