OpenAI introduces Ultrafast GPT-5.6 Sol mode

OpenAI announced the launch of the Ultra-Fast Processing Mode for the GPT-5.6 Sol model, which enables the generation of up to 750 tokens per second at speeds up to 14 times faster than the standard system, while retaining the full capabilities of the model without the need for a smaller version. This new mode is based on the Cerebras Wafer-Scale Engine architecture, utilizing 44GB of SRAM, which reduces data transfer bottlenecks and enhances processing speed. Experiments have shown that this mode outperforms the Fable 5 model by about 11 times and the Anthropic Opus 4.8 model by 5 times, with the ability to process 2,500 questions in approximately 11 hours, compared to 78 hours for another model, all while maintaining high accuracy. The Ultrafast mode is used for tasks requiring immediate responses such as emergency support, financial research, customer service, data analysis, e-commerce, and programming, allowing for record analysis and accelerated searches and experiments. Currently, its availability is limited due to capacity constraints, and OpenAI is evaluating customer requests based on workload and available capacity.

articleDetails.fontSize.sizeحجم النص

تنويه: هذا ملخص تم إنشاؤه بواسطة الذكاء الاصطناعي

عرض المقال بالكامل