ON OCTOBER 15, ACCORDING TO BGR, WHICH WAS PUBLISHED YESTERDAY, OCTOBER 14GoogleOn October 22nd, the next generation will be released AI Models Gemini 3.0.MAJOR IMPROVEMENTS IN DELAY, COST AND CODING AND SIGNIFICANT ENHANCEMENT OF SVG IMAGE GENERATION PERFORMANCE。
After OpenAI released GPT-5 in early August, and Anthropic launched Claude Sonet 4.5 in late September, competition in artificial intelligence became hotter and hotterThe focus of the industry is on Google, another heavyweight player, and is expected to launch Gemini 3.0 on October 22nd。
1AI refers to Boven's presentation that the core evidence of this explosion is a screenshot disseminated on social platforms such as Reddit and X, which appears to be a “significant milestone” planning form within Google。
The table refers explicitly to Gemini 3.0 planned for 22 October, and the contents of the document indicate that Google may have already started the Gemini 3.0 internal test earlier this August。
According to current sporadic information, Gemini 3.0 is expected to surpass the current top model Gemini 2.5 Pro at multiple dimensions. First, the new model is expected to optimize delays and cost control, which means faster, more economical AI services. Second, anecdotally, it is said to enhance its coding capabilities and to expand its intelligent body function to enable it to perform complex tasks more autonomously than merely exist as a chat robot。
More specific performance enhancements are reflected in specific mandates. A comparative map published by users on the Internet shows that Gemini 3.0 performed much better in generating SVG (Scalable Vector Graphics) than Gemini 2.5 and other models, with significant leaps in image detail and accuracy, although the authenticity of these images remains to be verified。
In addition, there are media speculations that Google may simultaneously upgrade its video generation tool to Veo 3.1 and launch a light quantification model of Nano Banana driven by Gemini 3.0 Pro。