Google has postponed the release of its highly anticipated Gemini 3.5 Pro artificial intelligence model, missing its initial June 2026 target. The delay comes as the company works to address performance shortcomings, specifically in the model's ability to generate and debug computer code. While Google initially aimed for a June launch following its May developer conference, the flagship model remains in testing with partners as engineers strive to meet internal quality benchmarks.
Alphabet shares saw a decline of approximately 4% to 5% following reports of the delay, reflecting investor sensitivity to the company's progress in the competitive AI landscape. The setback highlights the challenges Google faces in balancing rapid innovation with the rigorous standards required for its vast product ecosystem, which includes Search, Maps, and YouTube. Internal reports suggest that recent efforts to improve the model's coding capabilities through updated training data have not yet yielded the desired results.
Despite the delay, Google maintains that it is shipping updates across a wide range of models while prioritizing cost-effectiveness for its customers. The company confirmed it is actively testing Gemini 3.5 Pro and an upgraded version of its Flash model with partners. Furthermore, Google is engaging with the U.S. government regarding model testing and safety frameworks, a process that has also influenced the release timelines of other major AI developers in the industry.
For users and developers, the delay means waiting longer for the next iteration of Google's most powerful AI. While the company has not provided a firm new release date, the focus remains on refining the model's reasoning and coding performance. As the AI sector continues to evolve, the ability to deliver reliable, high-performing tools remains a critical factor for Google's competitive standing against rivals like OpenAI and Anthropic.