[Baidu Search AI Assistant Fully Launches Speed Model, Significantly Enhancing Search Result Generation Speed] On August 29, Baidu Search AI Assistant fully launched its speed model, leveraging self-developed technology to significantly enhance the speed of search result generation. Compared to DeepSeek V3.1, while maintaining the same effectiveness, the generation speed is five times faster, and the invocation cost is only 70%. This technology focuses on the inference phase within large models, comprehensively optimizing generation speed, invocation cost, and user experience. With the support of this technology, Baidu AI Assistant's first token return time is reduced by 39%, and response articulation speed is accelerated by 500%, making it the fastest in the industry. (People's Financial News)