Xiaomi is making a sweeping pivot toward artificial intelligence, committing more than 60 billion yuan (US$8.8 billion) over the next three years to develop frontier AI models — a move the Beijing-based company says is essential to keeping its smartphones and electric vehicles competitive in an AI-driven market.

MiMo-V2.5-Pro tops open-source AI benchmark

Xiaomi's latest AI model, MiMo-V2.5-Pro, introduced last month, was ranked by third-party benchmark platform Artificial Analysis as the world's leading open-source model for agentic capabilities — the ability of an AI system to autonomously execute complex, multistep tasks. On the same benchmark, the model placed second and third in overall intelligence and coding evaluations respectively, trailing only DeepSeek and Moonshot AI among open-source systems.

The result is notable given that Xiaomi only began releasing open-source AI systems approximately one year ago, placing it well ahead of many established technology peers on the agentic performance curve.

Why it matters

Hayden Hou, principal analyst at research firm Canalys, said the strategic logic is straightforward: 'AI is the most critical factor for all tech consumer goods companies.' As AI capabilities become increasingly embedded in consumer hardware — from camera processing to in-car autonomy — device makers without proprietary models risk ceding differentiation to software-first rivals.

For Xiaomi, which has built one of the world's largest consumer hardware ecosystems spanning smartphones, smart home devices, and now electric vehicles, owning the AI layer could determine whether its products command premium pricing or are commoditised.

The competitive backdrop

Xiaomi's AI push unfolds against intensifying competition in China's technology sector, where DeepSeek and Moonshot AI have set a high bar for open-source model performance. The company is also developing its own in-house chip, the XRing O1, signalling ambitions to control silicon as well as software — a vertical integration strategy that mirrors approaches taken by Apple in the West.

Analysts noted that combining proprietary chips with proprietary AI models could give Xiaomi meaningful cost and performance advantages in both its smartphone and EV divisions over the medium term.

What's next

With the 60 billion yuan commitment now on record, attention will turn to how quickly Xiaomi can translate benchmark wins into tangible product differentiation — particularly as its EV lineup scales and competition from domestic rivals intensifies. The pace of XRing O1 deployment and subsequent AI model releases will be the clearest indicators of whether the company's frontier AI ambitions are on track.