Google Delays Gemini 3.5 Pro Launch to July

June 24, 2026 - 17:21

Google's next major artificial intelligence model, Gemini 3.5 Pro, will not arrive in the second quarter as originally planned. The company now expects to release it in July, according to internal sources familiar with the matter.

The delay comes as Google works to improve the model's performance across several key areas, including reasoning, coding, and multimodal capabilities. Early internal testing reportedly revealed inconsistencies in how the model handled complex, multi-step tasks, prompting engineers to push back the launch timeline.

Gemini 3.5 Pro was initially expected to debut in late spring, building on the foundation laid by its predecessor, Gemini 2.5 Pro. That model, released earlier this year, was praised for its strong performance in math, science, and long-context understanding. The 3.5 version was supposed to take those abilities further, but the extra development time suggests the gap between promise and delivery is wider than Google anticipated.

The company has not made an official public announcement about the delay. Instead, the revised timeline has been communicated internally to teams working on product integration and testing. Google is reportedly using the additional weeks to refine the model's safety guardrails and reduce instances of hallucination, a persistent problem across the AI industry.

Competitors have not stood still. OpenAI recently updated its GPT-4o model, and Anthropic released Claude 3.5 Sonnet with improved speed and accuracy. Google's delay means it will enter a more crowded and demanding market when Gemini 3.5 Pro finally arrives.

For now, developers and enterprise customers who were expecting the model in June will have to wait a bit longer. Whether the extra time results in a genuinely superior product remains to be seen.