Recently, media reports have indicated that WeChat is conducting a gray-scale test of its AI question-answering feature. Some users can activate this function by typing questions into the WeChat search bar. The AI can generate a response of about 400 words within approximately 20 seconds and provide five reference materials from WeChat public accounts.
In response, the WeChat team stated that the purpose of this testing is to optimize the search experience for users and enhance search efficiency. This feature integrates technical resources from within Tencent, including the Hunyuan large model.
It is reported that the Tencent Hunyuan large model is a large-scale language model developed by Tencent, based on the Transformer neural network architecture, with a trillion parameter scale. It possesses powerful Chinese creative abilities, logical reasoning capabilities in complex contexts, and reliable task execution abilities.