BIDU(US:BIDU) Zheng Quan Shi Bao Wang·2025-11-13 11:33Core Insights - The Baidu World Conference held on November 13 showcased the company's advancements in artificial intelligence (AI), emphasizing that internalizing AI capabilities transforms it from a cost to a productivity driver, fostering efficiency and innovation across industries [1] Group 1: AI Technology Developments - Baidu released the Wenxin large model 5.0, a unified native multimodal model that enhances understanding, generation, logical reasoning, creative writing, and multimodal instruction adherence [3] - The new Kunlun chips M100 and M300 were introduced, designed for large-scale inference and ultra-large multimodal training, with plans for market release in 2026 and 2027 respectively [3] - The Tianchi 256 and Tianchi 512 super nodes were announced, expected to be available in 2026, with the Tianchi 512 capable of training trillion-parameter models, offering higher efficiency and lower-cost AI computing support [3] Group 2: AI Applications and Performance - Baidu's autonomous driving platform "Luobo Kuaipao" reported coverage in 22 cities globally, with over 140 million kilometers driven without human intervention and more than 17 million completed orders, averaging 250,000 fully autonomous orders weekly [3] - The integration of AI into Baidu's core products has led to a complete AI-driven reconstruction of its search engine, with 70% of search results now featuring rich media content [4] - The "Huibo Xing" digital human technology was utilized by 83% of live stream hosts during the 2025 "Double 11" event, resulting in a 119% increase in the number of live streams and a 91% rise in gross merchandise volume (GMV) [4] Group 3: Global Expansion of AI Solutions - Baidu's AI services are accelerating internationalization, with products like Huibo Xing digital human, GenFlow, and MeDo making inroads into overseas markets [5] - The company aims to internalize AI as a native capability to drive productivity revolutions across various sectors, with ongoing investments in advanced models and key technologies [5]