Google prepares Gemini 4 release in bid to catch up in AI race

Gemini 4 could serve as a sentiment driver for Alphabet shareholders. A strong model would indicate that the firm's AI investments are yielding results, whereas a poor debut might reignite concerns about its competitive standing. Investors across the AI sector may recalibrate their positions as they compare the new model to others. With no specific launch date announced, the immediate market impact is expected to be limited to anticipation rather than a price adjustment. Once the model is released, initial performance figures will carry more weight than the launch event itself.

---

AI in the news:

• Claude from Anthropic identified a CRISPR-like enzyme system after scanning 200,000 enzymes
• A government site was hacked by an AI agent for the first time, raising containment concerns

---

Google is pushing to launch Gemini 4 quickly and refine it post-release, relying on speed to recapture lost ground against Anthropic and OpenAI.

Summary:

• Google is readying its next major AI model, Gemini 4, for release.
• The model is in the early post-training phase, where the base model undergoes fine-tuning.
• Google aims to launch well before the end of the year and release an early version as soon as possible.
• The company plans rapid iteration after the initial release.
• Google is perceived as having fallen behind Anthropic and OpenAI in the AI model race.
• No specific release date has been provided.

According to remarks from Google DeepMind's Kavukcuoglu reported by The Information on Wednesday, Google is getting ready to roll out Gemini 4, its next leading AI model, with a target of releasing it well before year's end.

Kavukcuoglu noted that the model is in the early post-training phase, where the base model undergoes fine-tuning. He stated that Google aims to release an early version from this phase as quickly as possible, and that the outcomes so far have been promising. Following that launch, the company intends to continue enhancing the model at a fast pace.

The release is under scrutiny as Google is perceived to have fallen behind competitors Anthropic and OpenAI in developing top AI models. A successful Gemini 4 would bolster Google's argument that its significant AI spending is yielding marketable products, whereas a lackluster launch would amplify worries about its ability to stay competitive.

The strategy outlined by Kavukcuoglu—releasing an early version and then improving it via regular updates—mirrors the fast pace at which top AI developers now operate. New model launches are a recurring source of competition, with each quickly assessed against competitors by developers, businesses, and investors.

No precise release date has been announced, and the schedule depends on post-training progress. Nonetheless, the remarks indicate Google's eagerness to deliver its next-gen model to users soon, implying that the last part of the year may see renewed rivalry among leading AI firms. For shareholders, how Gemini 4 is received will likely be seen as a measure of whether Google can narrow the gap with its main competitors.