Tencent Unveils Hy4 Preview with 770B Parameters and 1M Context
Tencent released and open-sourced its new large language model Hy4 preview on August 28. The model has 770 billion total parameters, 49 billion active parameters, and a context length exceeding 1 million tokens. It targets real-world productivity tasks in coding, office work, and science. Tencent shares rose over 2% in afternoon trading following the announcement.
On the afternoon of August 28, Tencent released and open-sourced its next-generation large language model, Hy4 preview. The model has 770 billion total parameters, 49 billion active parameters, and a context length exceeding 1 million tokens. It is primarily designed for real-world productivity tasks such as coding, office work, and scientific applications. Tencent said that Hy4 preview represents significant expansion in model size, context length, and data scale, with combined advances in pre-training and post-training delivering improved intelligence.
Since Tencent Hunyuan rebuilt its AI infrastructure in February, it has iterated a major version every two months on average. Following the release of Hunyuan Hy3 on July 6, Hy4 preview arrived in less than two months, accelerating the pace of major version updates. Hy4 preview's total parameters exceed Hy3's 295 billion, and its active parameters are more than double those of Hy3.
Hy4 preview is now open-sourced and simultaneously available in Tencent products including WorkBuddy, CodeBuddy (domestic and international versions), Yuanbao, and ima. Users can experience the model directly through these products or access it via API through Tencent Cloud Tokenhub and OpenRouter. Pricing is set at RMB 6 per million input tokens, RMB 18 per million output tokens, and RMB 0.3 per million tokens for cache hits. In the afternoon trading session on August 28, Tencent's share price rose more than 2%.