Google confirms development of Gemini 4 with most ambitious pre-training run yet
> cd .. / HUB_EDITORIALE
News

Google confirms development of Gemini 4 with most ambitious pre-training run yet

[2026-07-27] Author: Ing. Pietro Maiorana
> share
Zenithby Meteora Web The operating system for your business. Social, clients, bookings and invoices in one platform. Gyms, barbers, professionals. Discover Zenith Free demo · no card

During Alphabet’s Q2 2026 earnings call, CEO Sundar Pichai provided concrete updates on the future of Google’s AI models. Following the delay of Gemini 3.5 Pro, the company has teased the next-generation Gemini 4, describing it as the most ambitious pre-training effort to date. The announcement comes amid fierce competition from OpenAI, Anthropic, and others.

Gemini 4 pre-training aims to push beyond all previous limits

Pichai stated that Google is allocating massive compute resources to ensure Gemini 4 remains competitive at the frontier of artificial intelligence. He said: 'We are training Gemini 4 with a very ambitious approach', emphasizing that much larger base models will be required to stay ahead. This marks a strategic shift from the Gemini 3.x series, which included models like Gemini 3.6 Flash and 3.5 Flash-Lite. Google now targets a generational leap, with a focus on agentic coding and complex reasoning.

Sponsored Protocol

The delay of Gemini 3.5 Pro paves the way for a fall release

Originally scheduled for June, Gemini 3.5 Pro’s launch was postponed. However, Google continues to iterate on Flash models, aiming for a near-monthly release cadence. Based on past patterns, Gemini 4 is expected to launch between November and December 2026. Pichai expressed confidence in internal progress, promising that users will be pleased when the model is released.

Meanwhile, Google is navigating a shifting landscape. For instance, Nvidia and Microsoft launched the Open Secure AI Alliance, excluding OpenAI, Google, and Anthropic, signaling fragmentation in the AI ecosystem. This move could impact market dynamics, but Google appears determined to forge its own path with Gemini 4.

Addressing agentic coding challenges

Pichai acknowledged that coding, particularly agentic coding, remains an area for improvement. Google is dedicating significant resources to enhance model capabilities in this domain, aiming to compete with offerings from OpenAI. The integration of agentic coding features into Flash models demonstrates a commitment to incremental progress, while Gemini 4 promises a fundamentally new architecture.

Sponsored Protocol

For more on recent Android updates, check out how to enable automatic document backup on Android, another noteworthy tech development.

In summary, Google is betting heavily on Gemini 4 to reclaim AI leadership. With ambitious pre-training and a clear strategic vision, the company aims to outperform competitors and set new standards. For more details on the Gemini model, visit the Wikipedia page for Gemini.

Source: https://9to5google.com/2026/07/26/google-gemini-4-teases

> share
Ing. Pietro Maiorana

> AUTHOR_EXTRACTED

Ing. Pietro Maiorana

Ingegnere informatico e co-fondatore di Meteora Web, CMO dell'agenzia. Esperto di marketing digitale, social media, advertising, copywriting e SEO.
[ Read Full Dossier ]

> METEORA_WEB // DIGITAL AGENCY

We build the digital presence your business deserves.

Websites, social media, online advertising, e-commerce and high-performance hosting, engineered with method by computer engineers in Sciacca, for all of Italy.

> MW_JOURNAL

> READ_ALL()