Chinese tech giant Tencent has officially launched its T1 reasoning model, marking a significant step in the country’s rapidly expanding artificial intelligence sector.
The upgraded T1 model boasts faster response times and improved capabilities for processing long-form text, the company announced in a post on its official WeChat account. Tencent highlighted that T1 maintains “clear content logic and neat text formatting,” with an “extremely low” hallucination rate—an industry challenge where AI generates misleading or false information.
The launch comes as Tencent intensifies its AI investments, following the introduction of its Turbo S foundational language model late last month. Turbo S is designed to process queries at a higher speed than DeepSeek’s R1 model, a key competitor in China’s AI landscape.
A performance comparison chart shared by Tencent suggests that T1 outperforms DeepSeek R1 on certain knowledge and reasoning benchmarks.
Previously, Tencent had offered a preview version of T1 through platforms such as its AI assistant application, Yuanbao. The official release strengthens Tencent’s push in the AI space as competition heats up with rivals like DeepSeek, which has developed models that deliver performance comparable to or better than Western AI systems at significantly lower costs.
The move aligns with Tencent’s broader AI strategy, which includes increased capital expenditure in 2025. The company has already ramped up aggressive AI investments throughout 2024, signaling its commitment to maintaining a competitive edge in China’s evolving AI market.
