Google Confirmed the Development Process of the New Generation Artificial Intelligence Model Gemini 4

Google confirmed the development process of the new generation artificial intelligence model Gemini 4. Sundar Pichai's statements and the details of the model are in our news. Google started to share the first clues about the new generation artificial intelligence models after the delay of the Gemini 3.5 Pro version, which was expected in June. The company has officially confirmed its work on Gemini 4 with the announcements of Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber.

Speaking on Alphabet's second quarter 2026 earnings call, CEO Sundar Pichai stated that Gemini 4 is the company's most ambitious pre-training process to date. Pichai emphasized that the competition in the field of artificial intelligence is very dynamic and Google is determined to be at the forefront of this race. Sundar Pichai underlined that the Gemini 4 model needs a much larger base model to remain competitive.

In its artificial intelligence development processes, the company gives priority to the allocation of Tensor Processing Units (TPU) in order to compete in frontier technologies. Google aims to improve especially in areas such as coding and agent-based coding. Pichai expressed his excitement about the progress made within the company with Gemini 4 and expressed his belief that users will be satisfied when the model is launched.

Considering past product releases, Gemini 4 is expected to be released in November or December 2026. The company aims to raise the bar in artificial intelligence technologies with this new generation model. Google also announced that development on existing Gemini 3.x Flash models will continue. The company is particularly focused on increasing agent-based coding capabilities.