Google to introduce new Gemini AI model
Google is preparing to release Gemini 4, a new flagship version of its artificial intelligence model. The company expects to make it available significantly earlier than the end of 2026.
Abdulrahman Sadikhli1 min read
Google is preparing to release Gemini 4, a new flagship version of its artificial intelligence model. The company expects to introduce it significantly earlier than the end of 2026. This was stated by Koray Kavukcuoglu, Google DeepMind's chief technology officer.
According to him, the company plans to make an early version of the model, which has completed the post-training stage, available as soon as possible, citing highly promising results from its development.
Gemini 4 is currently undergoing post-training — a stage in which developers improve the model's performance, reasoning capabilities and reliability through additional training and evaluation.
Competition in the AI sector
Google is accelerating work on the project amid intensifying competition in artificial intelligence. OpenAI, Anthropic and other major technology companies continue to develop increasingly capable models, putting pressure on Google and requiring it to maintain its position in a rapidly changing market.
A specific release date has not been announced. However, company comments suggest that Gemini 4 could appear significantly earlier than the end of the year if the remaining testing and development proceed as planned.
Integration into Google products
The launch could be particularly important for Google because Gemini is being integrated into an increasing number of the company's products and services, including Search, Android and Workspace. The new flagship model could affect not only AI applications but also how millions of users interact with the Google ecosystem.
Source: azernews.az
