Google plans to release its next flagship AI model, Gemini 4, before the end of 2026. Koray Kavukcuoglu, a leader at the DeepMind division, confirmed that the model is currently in the final stages of development. The company intends to launch a preview version shortly after the training process concludes to accelerate its release cycle.
The company has not launched a new flagship AI model since the Gemini 3 series debuted in November 2025. During this period, competitors OpenAI and Anthropic released their own advanced models, GPT-6 and Mythos, respectively. These releases have placed Google in a position where it is currently trailing behind its main rivals in the artificial intelligence sector.
Google previously missed its internal targets, including the planned June 2026 launch of Gemini 3.5 Pro, which CEO Sundar Pichai had announced at the I/O conference. Kavukcuoglu explained that the company opted to prioritize the release of smaller, more efficient Flash models instead of the Pro version to maintain development momentum.
Despite the current competitive landscape, Google leadership remains confident in the company's ability to maintain a leading position in AI research and development. The upcoming release of Gemini 4 is intended to help the firm return to a more consistent schedule for deploying new artificial intelligence technologies.
