Gemini 3 attracts attention before release

VHO - Early tests show that Gemini 3 outperforms Gemini 2.5 and Sonnet 4.5, promising to elevate Google's AI in software development.

Google is moving closer to releasing Gemini 3, slated for October 9, after extensive A/B testing in AI Studio. Early reviews show the new model delivers impressive performance, especially on coding-related tasks.

According to the development and research community, Gemini 3 is not only a significant improvement over Gemini 2.5, but also shows a clear advantage over Sonnet 4.5 – Anthropic’s recently released product. One highlight is the ability to process and create complex SVGs, which is considered an important measure of coding and technical capabilities, opening up expectations for wide application in software development as well as many other technology fields.

Image source: TestingCatalog.

Notable updates coming with Gemini 3

Along with the launch of the Gemini 3, insiders believe that Google will also introduce the Veo 3.1 and a new Nano Banana model – which is said to be built on the Gemini 3 Pro platform instead of the previous Gemini Flash. This change could help raise the bar in both image creation and AI workflows, while asserting its competitive position against rivals such as OpenAI or Anthropic.

In addition, Google is testing a “My Stuff” feature in the Gemini interface, where users can easily access AI products they have created, similar to the photo library on ChatGPT. In addition, Google is also renaming the “apps” section to “connected apps”, and integrating more closely with other Google services, reducing manual operations but may limit users’ ability to control details.

Agent Mode and Browser Automation

Another highlight under development is Agent Mode, which allows AI to take control directly in the browser. Recent betas show signs of adding automation features, bringing Gemini closer to the capabilities of ChatGPT and Copilot. This would allow users to assign AI actions right in the context of browsing, greatly expanding its practical applications.

AI Studio and long-term orientation

For AI Studio, Google plans to add some minor interface changes, reflecting its efforts to standardize and refine the generative AI ecosystem. These improvements will primarily benefit developers, early adopters, and technical content teams. However, Google emphasizes that users should stay tuned for official announcements for full details on access and features.

Image source: TestingCatalog.

Google's milestone in the field of artificial intelligence

This October release week is considered one of the most important milestones for Google in asserting its leadership in the field of generative AI. With Gemini 3, the tech giant hopes to not only improve the experience of individual users but also reshape business workflows, creating a profound impact across the industry.

According to Testing Catalog

Source: https://baovanhoa.vn/nhip-song-so/gemini-3-gay-chu-y-truoc-gio-phat-hanh-172708.html