Chinese smartphone and electric vehicle maker Xiaomi on Friday unveiled a new reasoning artificial intelligence (AI) model developed in-house, underscoring the company’s ambition to integrate its hardware products with home-grown generative AI.
The open-source MiMo model has 7 billion parameters and outperformed OpenAI’s o1-mini and Alibaba Group Holding’s QwQ-32B-Preview, part of the Qwen series of models, in maths reasoning and coding, Xiaomi said in a statement. Alibaba owns the South China Morning Post.
The model was developed by Xiaomi’s specialised AI task force, known as Core, the company said.
Xiaomi’s stock price in Hong Kong gained 4.7 per cent in morning trading on Friday, while shares of Kingsoft Cloud Holdings – in which Xiaomi holds a 10 per cent stake and Xiaomi CEO Lei Jun holds 11 per cent – jumped 15.3 per cent.
The launch of the model aligns with earlier reports that Xiaomi has been building up its computing resources. According to a report by local media outlet Jiemian in December, Xiaomi bought about 10,000 graphics processing units to train its models.
Xiaomi’s AI ambitions were evident when the company made an offer to hire Luo Fuli, China’s AI “genius girl” from DeepSeek. Luo, a key contributor to the DeepSeek-V2 model, ultimately declined the offer.
