Google’s CEO announced that the Gemini 3.5 Pro is currently undergoing testing, and the company is already focused on developing the next iteration, Gemini 4, with its most ambitious pre-training run to date. This strategic push reflects Google’s commitment to rapidly advance its Gemini models, which is essential for maintaining its leadership in generative AI, as pre-training runs are vital internal milestones for major AI labs like Google.

Alphabet: Alphabet Inc. serves as the parent company of Google and oversees its core internet services alongside AI initiatives through subsidiaries like Google DeepMind. The company drives innovation in large language models and generative AI tools. In this announcement, Alphabet’s leadership is providing updates on the internal development status of its flagship Gemini AI lineup.
Gemini 4: Gemini 4 is the upcoming successor in Google’s Gemini AI model series, aimed at pushing boundaries in capability and efficiency. The project involves large-scale pre-training runs managed by dedicated engineering teams. Recent statements position it as a key focus in the company’s frontier AI efforts.
Gemini 3.5 Pro: Gemini 3.5 Pro is a high-performance large language model from Google’s AI division designed for advanced reasoning and multimodal tasks. It forms part of the iterative Gemini family that powers consumer and enterprise products. The news indicates the model remains in active testing ahead of wider deployment.

AI Roadmap: Google is advancing successive versions of its Gemini models in rapid succession to maintain leadership in generative AI.
Model Training: Pre-training runs for next-generation models represent core internal milestones at major AI labs including Google.