Status as of 1 July 2026: Gemini 3.5 Pro is not generally available. It is in limited preview for select Vertex AI enterprise customers only. Google has confirmed a July 2026 target but has not given a specific date. The delay is attributed to ongoing refinements to coding accuracy, token efficiency, and long-task performance based on early tester feedback.

What was expected and why it matters

Prediction markets assigned 50-55% odds to Gemini 3.5 Pro shipping by June 30 — roughly a coin flip. It did not. The model became the most-watched AI product deadline of Q2 because of two genuinely differentiated features:

  • 2-million-token context window — the largest context of any major model, doubling Claude Opus 4.8 and every current OpenAI offering. This matters specifically for processing entire legal contracts, long project histories, or large codebases in a single session without chunking.
  • Deep Think reasoning mode — extended reasoning similar to OpenAI's o3 thinking models but natively integrated with Google Workspace. For complex financial analysis, multi-step planning, or research tasks, this represents a genuine capability jump over Gemini Flash.

For UK businesses on Google Workspace, the native integration was the real draw — Gemini 3.5 Pro inside Gmail, Docs, and Sheets without switching tools. That integration remains unavailable for Pro specifically.

What is actually available right now

Three Google AI products are live today that many UK businesses underuse:

  • Gemini 2.0 Flash — available in Google AI Studio and via API. Fast, cheap, capable of most routine business tasks. Not Pro quality but sufficient for email drafting, summarisation, FAQ generation, and simple analysis.
  • Gemini in Google Workspace — the Gemini sidebar in Gmail and Docs on eligible Business and Enterprise plans. Drafts emails, summarises threads, generates document content. Live now.
  • Gemini Advanced — the consumer and Google One premium tier using a more capable model (currently 1.5 Pro-equivalent). Available at £19.99/month via Google One AI Premium.
The decision framework

Wait for Gemini 3.5 Pro only if: (a) you have a specific task that genuinely needs a 2M-token context window, or (b) you are evaluating Google as a primary AI platform and want to test Pro before committing. For everything else, start with what is available now — the habit-building matters more than the model version.

When to evaluate Gemini 3.5 Pro

Google's July confirmation gives a window but no date. The practical recommendation is to set an evaluation date of 15 July and check the Google AI Blog for a GA announcement. When Pro ships, the key questions for UK businesses are: Does the Workspace integration work for your specific plan tier? What is the pricing compared to Flash? Does Deep Think justify the cost for your actual use cases?

Operator actions

Today: If you use Google Workspace, enable the Gemini sidebar in Gmail (Settings → See all settings → Labs). Start using it for email drafting and thread summarisation. This is free on eligible plans and builds the habit before Pro arrives.
15 July: Check for Gemini 3.5 Pro GA announcement. If shipped, evaluate using the context window on your longest/most complex document type.
Do not: Delay other AI platform decisions — Claude Sonnet 5 and GPT-5.5 are both available now at competitive pricing.