Tencent increases Hy4 model capacity following high user demand
Tencent has scaled the inference capacity of its Hy4 artificial intelligence model by tenfold following capacity issues on its WorkBuddy platform.
Tencent announced on November 5 that it increased the inference capacity for its newly released Hy4 model by a factor of ten. The move follows the platform's initial launch, where high user volume on the WorkBuddy interface caused widespread usage limit interruptions.
Alongside the infrastructure expansion, the company increased free usage allowances for users. Tencent indicated that it remains in a process of reallocating computational resources to stabilize performance levels as demand for the tool persists.
The Hy4 model, which debuted on November 4, introduces enhanced visual processing capabilities and expanded training datasets compared to its predecessor. Industry data from Artificial Analysis places the model second among domestic Chinese competitors, while it holds a top position in benchmarks for open-weight agents. With an estimated operational cost of $0.78 per task, the model is designed to handle complex instruction following and long-context processing.