Tencent releases Turbo AI model it says is faster than DeepSeek
Sign up now: Get ST's newsletters delivered to your inbox
Tencent's Turbo S AI model is designed to respond as instantly as possible, distinguishing itself from the deep reasoning approach of DeepSeek’s chatbot.
PHOTO: BLOOMBERG
HONG KONG – Tencent Holdings joined the ranks of companies measuring themselves against DeepSeek by touting how much faster its latest artificial intelligence (AI) model is.
The Hunyuan Turbo S AI model, released on Feb 27, is designed to respond as instantly as possible, distinguishing itself from the deep reasoning approach of DeepSeek’s eponymous chatbot, Tencent said. The deployment cost of Turbo S has also been greatly reduced, the company added.
Tencent’s new release comes a month after Alibaba Group Holding benchmarked its latest Qwen AI model against DeepSeek, showing the rapid ascent to industrywide influence that the small start-up achieved since the beginning of 2025.
Tencent also shared figures showing its Hunyuan Turbo S was competitive against DeepSeek’s V3 model in commonly used AI tests.
Tencent, which alongside Alibaba is one of China’s biggest cloud infrastructure providers, is integrating DeepSeek’s technology into its WeChat platform in a move that triggered a jump in its share price.
DeepSeek has also been widely embraced by government entities across China, including in Hong Kong, where various departments are now using a ChatGPT-style tool based on its technology.
To address surges in demand, DeepSeek this week slashed pricing to access its application programming interface during off-peak hours.
The 20-month-old start-up, which emerged out of a quantitative hedge fund, has upended the global AI sector and revived confidence in the ability of Chinese tech companies to innovate. BLOOMBERG
