Gemini 4 launch is near, says new Google DeepMind chief
What's the story
Google DeepMind's new head, Koray Kavukcuoglu, has revealed that the company is close to launching its long-awaited Gemini 4 model. The announcement comes during his first media interaction in the role. Speaking with The Information, Kavukcuoglu said that Gemini 4 is currently being refined and Google plans to launch it "much earlier" than the end of this year.
Anticipation
Kavukcuoglu's vision for Gemini 4
Kavukcuoglu expressed his eagerness to release an early post-training output of Gemini 4.
He said, "Our intention is to, like, as soon as possible, to release an early post-training output because we see the results and we are excited."
The statement indicates Google's commitment to fast-paced iterations and a more consistent AI release timeline after DeepMind's leadership changes.
Market positioning
Gemini lagging behind competitors' models
Since the launch of the Gemini 3 series in November 2025, Google has not released a new flagship AI model.
In the same time frame, OpenAI's GPT-6 and Anthropic's Mythos series models have been launched.
Both these models outperform Gemini 3, giving their developers a significant advantage in the ongoing AI race.
Strategic shift
Google's approach to AI model releases
Google CEO Sundar Pichai had announced a major Gemini 3.5 Pro update would be coming in June, but it was never rolled out.
Kavukcuoglu explained this decision by saying the company "took a little bit of a step back" to focus on faster, less powerful Flash models instead.
Despite these changes, he remains confident about Google's position in AI development saying "In my mind, it's a certainty that we are always gonna be at the frontier."