Recently, the Apple China App Store's free app rankings saw a significant shift. Tencent's Yuanbao app surged, surpassing Doubao to claim the second spot. DeepSeek remains firmly in the top position. This reflects the intense market competition and the rapid development of China's large language model (LLM) industry.
Tencent Yuanbao, a productivity-focused app built on Tencent's HunYuan large language model, officially launched in May 2024. Initially, its marketing was limited, resulting in a significant user gap compared to competitors. According to QuestMobile, as of December last year, Yuanbao had only 2.11 million monthly active users, ranking eighth. However, since its transfer from Tencent's Technology Engineering Business Group to the Cloud and Smart Industry Business Group in January, it has undergone rapid iteration.
On February 13th, Tencent Yuanbao integrated the full-fledged DeepSeek-R1, supporting both HunYuan and DeepSeek LLMs. Just four days later, its self-developed deep thinking model, "HunYuan T1," was launched in a gray-scale release. With the gray-scale testing of WeChat Search's AI features, a massive influx of users overwhelmed Yuanbao, prompting Tencent to deploy it to support WeChat Search.
On February 21st, Yuanbao added a DeepSeek feature for image recognition and analysis. Users can send any image, and Yuanbao will provide an interpretation based on its content. Currently, all models within Yuanbao support image understanding and file parsing, significantly enhancing user experience.
Meanwhile, ByteDance's Doubao app also performs well in the LLM field, boasting 75.23 million monthly active users in December, making it one of China's most popular AI-native applications. However, Yuanbao's rapid rise intensifies market competition. Major manufacturers are enhancing product performance and user experience by integrating open-source or self-developed models to compete for market share.
It's worth noting that DeepSeek's rapid growth and the competitive pressure from major players pose significant challenges to smaller AI startups. These companies need to find unique development paths to navigate the increasingly fierce market competition.
