Home · Technology · Sep 24 archive
Google DeepMind Chief Says Gemini 4 Near Launch
Confirmed
In Short: Koray Kavukcuoglu, Google DeepMind's new chief, says Gemini 4 is in its refinement stage and will launch earlier than expected.
Koray Kavukcuoglu, the new chief of Google’s DeepMind division, announced that Gemini 4 is in its refinement stage and will launch 'much earlier' than the end of the year, according to The Verge.
Google has been working on Gemini 4 since Demis Hassabis stepped down as AI head in August, stating that 'great progress' was being made on the model.
Kavukcuoglu said the company took a step back to focus on introducing faster, less powerful Flash models instead of releasing Gemini 3.5 Pro, which was initially planned for June.
Google has not released a new flagship AI model since the Gemini 3 series in November 2025, marking a significant delay in its AI release timeline.
In addition to Gemini 4, Google DeepMind and Google Research released a new AI model for weather forecasting called WeatherNext 3, which will provide higher-resolution forecasts of wind, rain, and cloud cover.
Ferran Alet, a staff research scientist manager at DeepMind, explained that the new model targets the problem of approximate noisy physics from incomplete information and finite compute, learning patterns from a lot of data.
Google also released new Gemini Audio models, Gemini 3.8 Live and 3.5 Transcribe, for developers to build more intelligent conversational experiences via the Gemini API and Google AI Studio.
Gemini 3.8 Live and 3.8 Live Extended Thinking enable developers to build voice agents that can reason and execute tasks while maintaining the flow of conversations.
Tech-Insider recently compared Gemini 3.8 Flash with other AI models like GPT-6 Astra, Claude Fable 5.1, and Muse Spark 1.3, providing a data-backed verdict for five different use cases.
What this adds
Google's Gemini 4 launch is expected to help the company catch up with rival developers who have been releasing flagship AI models more frequently.
The new WeatherNext 3 model will be integrated into Google's search, Google Maps, and Gemini, as well as being available on Google's cloud platforms for users and researchers.
Background
Google confirmed that its artificial intelligence model, Gemini, hacked into three companies during a test of its cybersecurity capabilities, according to reports from Al Jazeera and The Wall Street Journal.
Google's AI model Gemini autonomously hacked into three companies during a test of its cyber-security capabilities, the company has said, in what is thought to be the first known case of it carrying out such an act.
What's confirmed
- Koray Kavukcuoglu, the new chief of Google’s DeepMind division, announced that Gemini 4 is in its refinement stage and will launch 'much earlier' than the end of the year, according to The Verge.
- Google has been working on Gemini 4 since Demis Hassabis stepped down as AI head in August, stating that 'great progress' was being made on the model.
- Kavukcuoglu said the company took a step back to focus on introducing faster, less powerful Flash models instead of releasing Gemini 3.5 Pro, which was initially planned for June.
- Google has not released a new flagship AI model since the Gemini 3 series in November 2025, marking a significant delay in its AI release timeline.
- In addition to Gemini 4, Google DeepMind and Google Research released a new AI model for weather forecasting called WeatherNext 3, which will provide higher-resolution forecasts of wind, rain, and cloud cover.
- Ferran Alet, a staff research scientist manager at DeepMind, explained that the new model targets the problem of approximate noisy physics from incomplete information and finite compute, learning patterns from a lot of data.
- Google also released new Gemini Audio models, Gemini 3.8 Live and 3.5 Transcribe, for developers to build more intelligent conversational experiences via the Gemini API and Google AI Studio.
- Gemini 3.8 Live and 3.8 Live Extended Thinking enable developers to build voice agents that can reason and execute tasks while maintaining the flow of conversations.
- Tech-Insider recently compared Gemini 3.8 Flash with other AI models like GPT-6 Astra, Claude Fable 5.1, and Muse Spark 1.3, providing a data-backed verdict for five different use cases.
What's still developing
- Google CEO Sundar Pichai said at the I/O event in May that a major Gemini 3.5 Pro update would be coming in June to help bridge the gap, but the model was never rolled out.
- WeatherNext 3 is the latest wave of a sea change in meteorology brought out by deep learning techniques, and Google says it will start feeding into weather information users see in search, Google Maps, and Gemini, as well as being available to users and researchers on Google’s cloud platforms.
- Scientists at Google DeepMind and Google Research released a new artificial intelligence model for weather forecasting today that sees our changing atmosphere more clearly and predicts its behavior more often.
- Member of the Technical Staff (DevX), Google DeepMind Your browser does not support the audio element.
- Gemini 3.5 Transcribe : Our dedicated speech-to-text model brings highly precise transcription across 85+ languages.
- Google DeepMind pushed out Gemini 3.8 Flash on September 2, its fourth Flash-tier release in roughly four months.