The situation as of 21 June: Google's Gemini 3.5 Pro is in limited enterprise preview on Vertex AI and has not yet launched publicly. Sundar Pichai set a June deadline at Google I/O. Nine calendar days remain. Prediction markets — where participants bet real money on outcomes — are pricing the probability of a June 30 public release at 50–55%. That is meaningful uncertainty for a flagship model that has been publicly committed to for over a month.
What Gemini 3.5 Pro is and why it matters
Gemini 3.5 Pro is positioned as Google's answer to the frontier reasoning gap that Gemini 3.5 Flash (launched May 2026) left open. Flash improved speed and cost but regressed on complex multi-step reasoning compared to Gemini 1.5 Pro. Gemini 3.5 Pro is designed to close that gap — and then some. The 2-million-token context window (double Flash's capacity) makes it the most capable model available for large-document analysis. Deep Think mode adds an explicit reasoning layer for hard multi-step problems. For UK businesses using Google Workspace with Gemini integration, this is a meaningful capability upgrade.
The model also replaces the Ultra tier in Google's product line — there will no longer be a separately branded "Gemini Ultra." Pro is the new top tier for general availability, with a potential research/enterprise-only tier above it.
Why the delay is notable
In the current competitive landscape, model launch windows are consequential. The month of June 2026 has had Anthropic (Fable 5/Mythos 5, 9 June), xAI (Grok 5 signals), and OpenAI (Realtime API GA on 19 June) all making significant moves. Google's public commitment to a June release for Gemini 3.5 Pro was partly a competitive positioning signal — "we are moving at the same pace." If the model slips to July, the narrative becomes "Google committed to June and missed," which is a different message in a market where timing is tracked closely.
For UK businesses using Google tools, the more practical concern is workflow planning. Several enterprise Gemini features are gated on Gemini 3.5 Pro's release. If you have been waiting for Pro's 2-million-token window to test a specific large-document workflow, you cannot plan that test for this week with confidence.
The prediction market signal
Prediction markets aggregate the views of many participants who put real financial stakes on outcomes, which historically makes them more accurate than analyst forecasts for near-term binary events. A 50–55% probability for a June 30 release means the market sees it as roughly a coin flip. The base rate implication: plan for the possibility that Gemini 3.5 Pro launches in the last week of June, but also have a contingency for a July release without being blindsided by either outcome.
This is not a story about Google being in trouble — Gemini 3.5 Flash is already live and performing well, and the 3.5 Pro limited preview is confirmed. It is a story about the complexity of shipping frontier models on a public timeline commitment when the full safety evaluation process takes the time it takes.
What Gemini 3.5 Pro means for UK Google Workspace users
If your business uses Google Workspace (Gmail, Docs, Sheets, Drive) with Gemini AI features enabled, Gemini 3.5 Pro will upgrade the underlying model powering those features at no additional subscription cost. The practical improvements will be most noticeable in tasks that currently produce poor results from Gemini — complex document summarisation, multi-step reasoning across large datasets, and nuanced writing where current Flash output misses important context.
For businesses not yet using Gemini in Workspace, the Pro launch is a natural moment to evaluate it — the model step-up makes the first experience significantly better than testing on the current Flash tier.
What UK operators should do now
Identify your Gemini 3.5 Pro test case now, not on launch day: Think about the one task in your business that current AI handles poorly due to context limits or reasoning quality. Write it down. When Pro launches — this week or next — that is your first test. You will get a faster, cleaner read on its value if you already know what you are testing.
Do not delay workflows waiting for Pro: If you need to make AI workflow decisions this week, make them based on currently available models. Gemini 3.5 Flash, Claude Sonnet 4.6, and GPT-4o are all fully available and production-ready. Do not let a 50/50 launch gamble hold up work that can proceed now.
Google Workspace users — check your Gemini settings: Log in to your Workspace admin console and confirm Gemini AI features are enabled. When Pro rolls out to Workspace, it will be an admin-controlled update — you want to make sure you are not on a delayed rollout tier.
Watch the Vertex AI release notes: The public launch of Gemini 3.5 Pro will appear in Google Cloud's Vertex AI release notes first, before the broader announcement. If you want early visibility, subscribe to those notes directly.