GOOGLE’S MOST POWERFUL MODEL IS COMING THIS MONTH — AND THE CROWD AT I/O GROANED WHEN SUNDAR WOULD NOT GIVE A DATE Sundar Pichai stood on stage at Google I/O on May 19 and confirmed that Gemini 3.5 Pro would be available next month, June 2026, but declined to give a specific date. The audience reportedly groaned. The reaction was understandable. Gemini 3.5 Flash had just launched to strong reviews, but it represented a deliberate tradeoff of reasoning accuracy for speed and cost. Pro is the version that fixes what Flash gave up, and developers who need it for production agentic workflows have been waiting. The product context matters. Gemini 3.5 Flash is already live at $1.50 per million input tokens and $9 per million output tokens, and Google is marketing it as frontier performance for agents and coding. The Pro tier is expected to land at two to five times Flash pricing, which would put it roughly in Claude Opus and GPT-5.5 territory on cost. On capability, Google’s internal positioning suggests Pro closes the reasoning gap that opened when Flash prioritized throughput over accuracy on multi-step tasks. For the enterprise AI coding market specifically, a capable Gemini 3.5 Pro would give Google Cloud customers a serious alternative to Claude Code and OpenAI Codex without leaving the Google ecosystem. That is the real strategic play. Not beating Anthropic on a benchmark table, but giving large enterprises already running Workspace, BigQuery, and Vertex AI a reason to keep AI inference inside the same vendor relationship. Anthropic and OpenAI both have strong coding agent stories. Google’s advantage is that 90% of the Fortune 500 already has a Google contract. Gemini 3.5 Pro is the model that turns that installed base into an AI workload pipeline. Keywords: Gemini 3.5 Pro, Google AI model release, Google I/O 2026, enterprise AI coding