New Google DeepMind Chief Says Gemini 4 Nearing Completion, Aims for Rapid Release

Google is reportedly putting the final touches on its highly anticipated Gemini 4 AI model, according to Koray Kavukcuoglu, the newly appointed head of Google’s DeepMind division. Speaking with The Information during his first media interview in the role, Kavukcuoglu revealed that Gemini 4 is currently in its refinement phase and that the company is targeting a launch “much earlier” than the end of the year.

“Our intention is to, as soon as possible, release an early post-training output because we see the results and we are excited,” Kavukcuoglu said. He added that Google plans to “continue the fast-paced iterations” moving forward, signaling a renewed commitment to accelerating its AI development timeline.

The announcement suggests a potential comeback for Google in the competitive AI landscape, particularly after a period of slower model releases. Former AI head Demis Hassabis stepped down in August, stating that significant progress was already underway with Gemini 4—a claim that appears to be materializing under Kavukcuoglu’s leadership.

Google’s last major AI release was the Gemini 3 series in November 2025. Since then, competitors like OpenAI and Anthropic have launched advanced models – GPT-6 and the Mythos series respectively – both of which have surpassed Gemini 3’s capabilities, widening the gap between Google and its rivals in the AI race.

At Google’s I/O developer conference in May, CEO Sundar Pichai hinted at a major Gemini 3.5 Pro update set to debut in June, but the model ultimately did not roll out as planned. Kavukcuoglu explained that the company shifted priorities toward developing faster, lightweight Flash models instead of pursuing the larger Gemini 3.5 Pro release.

When questioned about Google’s ability to keep pace with industry leaders in AI innovation, Kavukcuoglu confidently assured The Information that “in my mind, it’s a certainty that we are always gonna be at the frontier.”

Source link