Gemini 3.5 Pro reportedly slips to July
Gemini 3.5 Pro has reportedly slipped from June to July while Google gathers tester feedback. The delay matters because agentic coding has become the frontier-model battleground.
Gemini 3.5 Pro appears to be taking a little longer to leave the lab. Business Insider reports that Google has moved the expected launch from June to July while it gathers feedback from early testers and tunes the model for long-horizon agent work.
That is a small calendar slip, not a product failure. But it is meaningful because Google positioned the Gemini 3.5 family around exactly the category where OpenAI, Anthropic, xAI, Meta, Mistral, and AWS are fighting hardest: coding agents, tool use, and longer-running enterprise workflows. A frontier Pro model now has to be judged less by chatbot polish than by whether it can plan, call tools, conserve tokens, recover from errors, and keep costs predictable over multi-step jobs.
Google said at I/O that Gemini 3.5 Flash was available through the Gemini app, AI Mode in Search, Google Antigravity, the Gemini API in Google AI Studio, Android Studio, Gemini Enterprise Agent Platform, and Gemini Enterprise. In that same launch post, Google said it was working on Gemini 3.5 Pro, using it internally, and planned to roll it out the following month. Business Insider now reports that the target is July, citing a person familiar with the matter; Google declined to comment to the publication.
The practical read for builders is to keep shipping against 3.5 Flash for current Gemini 3.5 agent surfaces and wait for Google’s official Pro availability before hard-coding model IDs, pricing, quotas, or SLA assumptions. Business Insider says the upcoming Pro model has been available to some users through Antigravity and LMArena, and that Google is incorporating feedback from 3.5 Flash, including complaints that Flash could burn through tokens quickly. If that is the issue being tuned, a short delay may be rational: token discipline matters directly to enterprise agent cost and reliability.
For Google, the strategic upside is still distribution. When Pro arrives, it can flow through the same Gemini API, Vertex/enterprise, Search, Android, and Antigravity surfaces that already give Google a broad route to developers and enterprise customers. The risk is cadence: Anthropic and OpenAI have made agentic coding the default comparison set, so Google needs 3.5 Pro to feel not just smarter than Flash but operationally better for real work.
References: