Tencent Holdings said on Thursday companies could now use its large language artificial intelligence model “Hunyuan”.
The Chinese tech giant and owner of the WeChat social media platform conducted a demonstration before a live audience at a conference in Shenzhen, and said Hunyuan had become the foundation of more than 50 of its products and services.
“By July, there are more than 130 large language models in China,” Tencent’s vice president, Jiang Jie, said.
“A war of a hundred models has begun.” He added
After a number of Chinese IT companies, including Baidu Inc. and SenseTime Group, recently introduced their own AI models, Hunyuan made its debut.
Hunyuan was trained with more than 2 trillion tokens and had more than 100 billion parameters, according to Tencent, the most valuable internet business in China. These two criteria are frequently used to gauge the strength of AI models.
In 2020, the GPT-3 AI model from OpenAI had 175 billion parameters, and in 2023, the Llama 2 model from Meta had 70 billion parameters.
Tencent claimed that its model, which can communicate in both Chinese and English, was “better” than OpenAI’s ChatGPT in tasks including producing lengthy texts with a lot of words and resolving specific mathematical problems.
