Khaberni - Google announced today the launch of its new artificial intelligence model, Gemini 3.0, considering it the smartest system so far, featuring its ability to understand information with a depth and precision similar to human thinking.
This announcement comes days after speculations about the model's upgrade and its ability to excel over current models like "ChatGPT 5" and "Groq 4", according to indiatoday.
Multiple Capabilities
The company says that Gemini 3.0 combines multiple capabilities, including understanding multimedia information, analyzing long contexts, and Agentic Behaviour, which enables it to deal with complex problems in various fields, including science and mathematics, with a high degree of accuracy and reliability.
According to Google CEO Sundar Pichai, the new model is capable of capturing precise evidence in creative ideas and understanding the layers of complex problems, in addition to interpreting the context and intentions behind user requests better, which reduces the need for repeated guidance.
Pichai said in a blog post: "It is amazing to see how artificial intelligence has evolved over two years from just reading texts and images to understanding the full context and interacting intelligently."
Model Performance and Superiority Tests
Demis Hassabis, CEO of Google DeepMind, pointed out that Gemini 3.0 is capable of solving complex problems across a wide range of topics and provides smart and direct responses, avoiding clichés and flattery, in favor of delivering genuine and useful insights.
He added that the model excelled in academic and analytical tests, including Humanity's Last Exam, GPQA Diamond, and MathArena Apex, which measure the ability to solve mathematical and symbolic problems.
New Capabilities
The capabilities of Gemini 3.0 enable analysis and understanding of various types of content, including handwritten notes, long researches, and recorded lectures, and users can apply these capabilities in digitizing traditional recipes, summarizing complex academic papers, or deconstructing long videos into step-by-step explanations.
The model will also integrate with Google's artificial intelligence mode in search, where it can generate dynamic visual designs, simulations, and interactive results based on tools.
In addition to the main version, Google revealed a mode called Gemini 3 Deep Think, designed for complex analysis tasks, which showed superior performance over Gemini 3 Pro in internal tests, especially in tool use, code execution, and hypothesis verification.
Currently, this mode is not available for regular users and will be tested by carefully selected researchers, to be later released to Gemini Ultra subscribers after confirming its readiness.
The company has also expanded Gemini's capabilities for developers, particularly in programming, while maintaining complete user control over the system.
Google notes that more than 650 million users use Gemini systems monthly, reflecting the rapid expansion of artificial intelligence across all of the company's platforms.
