Alibaba Cloud, the cloud computing unit of Chinese tech heavyweight Alibaba Group Holding Ltd, unveiled on Friday its open-source 72 billion-parameter version of Tongyi Qianwen, the artificial intelligence-powered large language model and a potential rival to OpenAI's ChatGPT.

The LLM, called Qwen-72B, achieves a better performance over GPT-3.5 and GPT-4 in many tasks, with enhanced capabilities in understanding English and Chinese language, mathematics reasoning and coding.

The company has also released Qwen-1.8B and audio model Qwen-Audio. It has already open-sourced LLMs with 7 billion parameters and 14 billion parameters.

Zhou Jingren, CTO of Alibaba Cloud, said the open-source ecosystem is crucial for promoting the technological advancement and application of LLMs in China, adding they will continue to invest in open-sourced LLM, and aim to offer the most open LLM in the era of AI.

He added that the open-source LLM will help enterprises simplify the process of model training and deployment, lower threshold for the application of LLM and establish their own customized AI models at a faster pace.

fanfeifei@chinadaily.com.cn