Google Delays Release of Gemini 3.5 Pro AI Model Google’s biggest AI debut of the year did not arrive on schedule at its own stage, as the company used its I/O 2026 developer conference to promise more powerful models while asking developers to wait a few more weeks for its flagship Gemini 3.5 Pro.

Early expectations and the on‑stage letdown

In the run‑up to I/O, many in the industry expected Google to unveil its most advanced AI system yet. Instead, during Tuesday’s packed keynote, CEO Sundar Pichai told attendees that Gemini 3.5 Pro “wasn’t ready yet,” prompting audible groans from the crowd. Pichai asked developers to “give us until next month” for the release, positioning the delay as a short pause for a significantly improved model.

A strategic delay amid the AI coding race

Reporters covering the event argued that holding back the model reveals Google’s competitive calculus. One analysis framed it as Google “held back a huge new model, and it says a lot about the company’s standing in the AI coding race,” suggesting the company is likely optimizing Gemini 3.5 Pro specifically for coding tasks before launch. Rivals like Anthropic’s Claude Code and OpenAI’s Codex have recently surged, “scooping up developer mindshare and generating big revenue” from AI‑assisted programming tools.

To bridge the gap, Google promoted a smaller sibling model, Gemini 3.5 Flash, described as “faster and a lot cheaper, while being only slightly less powerful than the world’s current top models.” Flash now powers Antigravity, Google’s AI coding service, which is expected to generate valuable usage data to further refine Gemini 3.5 Pro.

Big promises, awkward pause

The delay came alongside sweeping AI ambitions. Coverage of the keynote noted that Pichai “brought big AI promises and one very awkward delay to I/O,” capturing the contrast between Google’s aggressive roadmap and the missing headline model. While new products like the Spark AI agent and advances in video generation underscored Google’s momentum, analysts also stressed that “Google isn’t quite as dominant in AI as it wants to be,” particularly in hot areas like AI coding.

Overall, the timeline from expectations to announcement shows a company trying to balance spectacle with strategy: delaying its most powerful model today in hopes of winning a longer AI race tomorrow.

Continue reading https://foxvector.com