Google preps Gemini 4 for a release ahead of schedule
Google DeepMind just pushed its flagship Gemini 4 model into the post-training phase, aiming for a release well before the end of the year. DeepMind lead Koray Kavukcuoglu noted that an early version could drop as soon as possible, as Google scraps Gemini 3.5 Pro to chase rival models from Anthropic and OpenAI.
Google is skipping the half-steps and gunning straight for its next flagship.
Gemini 4 enters the final stretch
Speaking at The Information's AI Agenda Live Summit, Google DeepMind head Koray Kavukcuoglu revealed that the Gemini 4 flagship AI model has officially entered post-training.
Post-training is the critical phase where researchers refine a raw base model, fine-tuning it to follow instructions and behave reliably.
Kavukcuoglu noted that he hopes to launch Gemini 4 "much earlier" than the end of the year, with an early output from the post-training phase arriving "as soon as possible."
Google previously scrapped its planned Gemini 3.5 Pro update. The search giant currently trails behind Anthropic's Mythos and OpenAI's GPT-6 in raw capability.
Why it matters
Scrapping Gemini 3.5 Pro to rush Gemini 4 out the door shows just how intense the pressure at the frontier has become. Google knows an incremental refresh wouldn't move the needle against GPT-6 and Mythos, so it is betting big on a leapfrog release.
For developers and enterprises, this means an earlier chance to test Google's next-generation intelligence, though rushing post-training always carries risks of rough edges. We think this aggressive timeline is Google's clearest signal yet that it refuses to stay in third place.
Sources (1)
- theinformation.com(opens in a new tab)Originaltheinformation.com