[Tencent Announces Progress on Hy3 Preview Language Model: Token Usage Increased Tenfold in Two Weeks] On May 7, Tencent Hunyuan released the latest data showing that since its launch, the token usage of the Hy3 preview model has continued to grow, with the total usage now exceeding ten times that of the previous-generation model, Hy2. Notably, token usage in code and intelligent agent-related scenarios has seen significant growth, with applications like Tencent's WorkBuddy/CodeBuddy and Qclaw experiencing an increase of over 16.5 times. On April 23, Tencent Hunyuan released and open-sourced the Hy3 preview language model. Hy3 preview is the first model launched by Tencent Hunyuan after a technical overhaul, utilizing a hybrid expert (MoE) architecture that integrates fast and slow thinking. The model boasts a total of 295 billion parameters, with 21 billion active parameters, and supports a long-context window of up to 256K. Hy3 preview has been fully integrated into multiple Tencent products, including Tencent Yuanbao, QQ Browser, WeChat Reading, and various other business scenarios. (China Securities Taurus)