Washington, DC – Google has announced the launch of its new artificial intelligence model, Gemini 3.1 Pro, making it available to users through its official app and to developers through the Google AI Studio and Vertex AI platforms. According to an official statement from the company, this move aims to support digital innovation and enhance the processing capabilities for complex programming and research tasks.
Advanced reasoning abilities
The new model features advanced reasoning capabilities that allow it to handle multi-step scientific and engineering problems more efficiently. According to the company’s technical test results, Gemini 3.1 Pro achieved a 77.1% success rate in the ARC-AGI-2 benchmark. It also outperformed previous versions, with significant improvements in response accuracy and error reduction.
Huge context window
ويعتمد النموذج على نافذة سياق تصل إلى مليون رمز مميز، ما يمكنه من تحليل مستندات ضخمة قد تصل إلى 1500 صفحة. ويمكنه أيضاً مراجعة مستودعات برمجية كاملة خلال وقت قصير. وتمنحه هذه الميزة قدرة متقدمة على فهم السياقات الطويلة والمعقدة. وهذا ما يمثل تطوراً مهماً في تطبيقات الذكاء الاصطناعي المؤسسية.
Smart multitasking agent
Gemini 3.1 Pro was also developed to function as an “intelligent agent” capable of performing autonomous tasks across third-party applications, providing accurate summaries, and offering technical solutions to problems. Additionally, it can generate interactive graphs and high-resolution SVG files.
According to Google, the new model is designed to combine high performance with efficient resource consumption. This makes it a strategic tool for developers and researchers, enhancing the company’s competitiveness in the rapidly growing AI technology market.
