Tencent, the Chinese technology powerhouse, unveiled the official version of its T1 reasoning model on Friday night, intensifying the competition within China’s burgeoning artificial intelligence industry.
The updated T1 model features quicker response times and improved capabilities for analyzing longer text documents, according to an announcement made on the company’s official WeChat account.
The post noted that T1 can maintain “clear content logic” and produce “neat and clean” text, while also boasting an “extremely low” rate of inaccuracies.
This launch comes at a time of escalating competition in China’s AI market, with DeepSeek having recently introduced models that claim to match or exceed the effectiveness of Western alternatives while offering significantly lower price points.
Prior to this official release, Tencent had offered a preview version of T1 through various platforms, including its AI assistant app, Yuanbao.
The official T1 model is built on Tencent’s Turbo S foundational language model, which was introduced just last month and is touted to process user queries more rapidly than DeepSeek’s R1 model.
A comparison chart shared in the post illustrated that Tencent’s T1 surpassed the DeepSeek R1 in several knowledge and reasoning benchmarks.
In recent months, Tencent has significantly increased its investments in AI technologies. On Thursday, the company revealed plans to boost capital expenditures in 2025, building on an already ambitious spending strategy for AI throughout 2024.
© Thomson Reuters 2025
