Google is new Ship 3 He began to test the artificial intelligence model on the AI Studio platform. This development is the company GPT-5 He shows that his move can now be very close. The screenshots shared by developers revealed that test versions are called “flash-reinforcement .. In addition, Google Cloud’s live broadcast on October 9 reinforces the likelihood of the first event to share official information about Gemini 3. This shows that Google is preparing to reshape the competition in the field of productive artificial intelligence.
With existing ship models Pro, Flash And Ultracan be used in multiple tasks such as text, visual and code production. The new version of the new version further improving these abilities In coding performance, In SVG drawings And In speed optimization It is expected to offer significant improvements. Developers reported that in the first results they achieved during the test process, code production is progressing more accurately and decisively. In addition, a significant quality increase in visual production processes is observed. Although the company did not confirm these results, the scale of the tests suggests that the launch is approaching.
According to benchmark leaks, Gemini 3 provided an increase in code up to 20 percent in code production compared to the previous version. In addition, a significant development is observed in the reasoning processes of the model. The system, which can solve more complex errors in coding tests, gives more consistent results in software production. In some sources, Gemini 3’s trial version offers higher performance values in similar scenarios with the GPT-5. Despite this, Google did not make an official statement confirming this data.
Google can introduce Gemini 3 at the event to be organized on 9 October
Google Cloud announced that it will organize a live broadcast on October 9 with the #Geminiatwork label. In this event, the details of Gemini 3 structure, usage areas and developer vehicles are expected to be shared. Also your model Workspace, Chrome And Android It is stated that sample demonstrations on how to integrate with Google products. The company plans to keep the Gemini family in the center to deliver productive artificial intelligence technology to a wider audience. This step can allow developers to perform faster and more reliable tests in AI Studio.
Some developers are in the trial versions of Gemini 3 browser -based task automation He claimed that their features were included. This feature can enable users to perform recurrent operations on the Internet with the help of artificial intelligence. In addition, new version labels such as “Gemini-3.0-pro” and “Gemini-3.0-Flash” were detected in Google’s internal code warehouse. These findings give the impression that the company has prepared separate versions for different hardware and use scenarios. Although these labels are said to be added for test purposes, it is thought that pre -launch developer environments have been expanded.
All these developments in the last week show that Google is about to announce the Gemini 3 model. The live broadcast of 9 October will be of great importance for the clarity of rumors in this direction. The company’s official statement may be decisive about both the future of the Gemini ecosystem and the direction of the competition with the GPT-5. Developers, with the possible promotion to be made at the event, when Gemini 3 will be available in general.