At MWC, Xiaomi demonstrated another path for the second half of the AI era.

When entering Fira Gran Via in Barcelona this year, you'll find that AI is the most prominent and widespread theme.

The theme of MWC (Mobile World Congress) 2026 is set as "The IQ Era", clearly pointing to "intelligence". Around this theme, operators, terminal manufacturers, chip and cloud service companies, and other manufacturers have started to showcase AI in more specific ways: stronger edge-side capabilities, faster real-time translation, smarter image processing, and more assistant-like interactions.

In this wave of AI-centered exhibitions, Xiaomi set the theme of its booth as "The New Wave of AI" and organized all the displays with a structure closer to people's daily lives: AI Today (intelligence in handheld devices) and AI Tomorrow (intelligence in travel and home spaces). From the system-level AI capabilities of the Xiaomi 17 series, such as HyperAI, the cross-device collaboration of HyperConnect, the intelligent travel space presented by the SU7 Ultra (with a lap record of 7 minutes and 04.957 seconds at the Nürburgring), to the exploration of home scenarios driven by the self-developed base large model MiMo, the core emphasized in the booth is a continuous experience from mobile phones to cars and then to homes.

This layout makes Xiaomi both similar to other manufacturers and shows a different path.

The similarity lies in that Xiaomi also provides a combination of hardware, systems, and AI capabilities, allowing visitors to see how AI improves the efficiency of imaging, translation, creation, and connection. The difference is that while many manufacturers still use single-piece hardware or individual chatbots to showcase the boundaries of AI, Xiaomi also demonstrates the way AI operates on a large scale in the new era - AI + hardware + ecosystem.

As of now, Xiaomi's "full ecosystem of people, cars, and homes" has connected over 1 billion devices, covering more than 200 product categories and reaching 95% of daily life scenarios. Its monthly active users of mobile phones globally exceed 740 million.

In the view of Lu Weibing, a partner and the president of Xiaomi Group, in the first half of the AI development, the competition is about model parameters and computing power. In the second half, the key is to combine model capabilities with specific scenarios, which is exactly Xiaomi's advantage: bringing AI out of the screen and into mobile phones, cars, and home spaces. "Making AI move from the virtual world to the physical world is Xiaomi's greatest advantage."

The AI competition is changing

If we divide the AI wave in the past three years into stages, the keywords in the first half were almost the same: model parameters, training costs, inference speed, and evaluation scores. Whoever had a larger model, stronger computing power, and a higher ranking on the list became the most intuitive competitive indicator in the industry. From the GPT series to various open-source models, technological breakthroughs constituted the most prominent narrative in this stage.

But the focus of the discussion has shifted.

At this year's MWC, this signal is even more obvious. As the intelligence of the models themselves gradually approaches the ceiling and the gap continues to narrow, the focus of the competition has become whether these capabilities can truly enter daily scenarios.

In the second half, three major transformations are taking place.

First, can AI be used across devices? If the excellent capabilities in mobile phones cannot be extended to tablets, computers, in-car systems, and home devices, the user experience will be fragmented. Today's users no longer live within a single screen but move between multiple terminals. Whether AI can seamlessly continue across different devices is an important standard to test its maturity.

Second, can AI understand real scenarios? From text conversations to image recognition and then to environmental perception, if AI only stays at the input-output level, it is difficult to truly change people's lifestyles. The truly meaningful progress is the ability to combine cameras, sensors, and behavioral data to make judgments about the environment and respond actively.

Third, can AI accompany users continuously? Instead of a one-time tool call, it should exist continuously in the daily life routine - from checking the mobile phone in the morning, during the commute, to returning home. Whether the experience is coherent and whether the intelligence is continuously online.

This industry trend will magnify the advantages of Chinese enterprises infinitely at this stage.

As the world's largest mobile Internet market, China has a mature digital payment and business closed-loop, a high penetration rate of intelligent terminals, and highly active user data. In such an environment, Chinese AI manufacturers can more easily polish and verify products in real scenarios.

In this transformation, Chinese enterprises as a whole have entered the leading echelon. A group of Chinese technology companies have gradually formed their own clear implementation directions: Alibaba promotes the implementation of AI based on its business and cloud ecosystem, ByteDance applies AI around its content production and distribution system, and Tencent embeds intelligence into social and collaboration scenarios.

Xiaomi takes a different route - relying on the "full ecosystem of people, cars, and homes", it integrates AI into mobile phones, cars, and home devices, making intelligence truly land in people's lives.

The former three are more inclined to the extension of the digital world, while Xiaomi is closer to starting from the physical world: mobile phones, cars, and home devices form a continuous life trajectory.

Integrate AI into terminals and then into life

All along, in the global AI competition landscape, Xiaomi doesn't rely on a single blockbuster model to win but combines model capabilities, hardware scale, and long-term accumulation in real scenarios to form a complete chain from underlying capabilities to terminal products and then to daily life.

The starting point of this chain is the construction of model capabilities.

In 2024, Xiaomi officially entered the field of base large models. By 2025, the MiMo language model, multimodal model, voice model, and embodied model were successively released and open-sourced. In February 2026, the hybrid sparse attention architecture HySparse for the Agent era was launched, seeking a new balance between efficiency and cost. As versions like Xiaomi MiMo-V2-Flash entered the global open-source first echelon in multiple comprehensive evaluations, Xiaomi is no longer just a player at the application layer but has begun to have its own model base.

More importantly, from the very beginning, MiMo was positioned as a unified source of capabilities for multiple terminals. As Lu Weibing shared in an interview: "Last year, a relatively good layout for AI large models was completed, including large language models, large voice models, and large multimodal models." Xiaomi's large models need to provide inference capabilities for the imaging and interaction of mobile phones, provide voice and perception support for in-car systems, and provide a basis for environmental understanding in home scenarios.

When the functions between different devices can be connected, the intelligent experience can continue on multiple terminals, thus reducing the sense of fragmentation for users.

The implementation of model capabilities cannot be separated from real terminals.

On Xiaomi's booth, this structure is clearly split. On one side of the entrance is the experience area centered around personal devices, where visitors can experience image enhancement, AI writing, real-time translation, and cross-device collaboration on-site. On the other side, there is a concentrated presentation of car and home scenarios, from the SU7 Ultra to the linkage demonstration of home spaces, where intelligence is integrated into a larger physical space.

In the AI Today section, Xiaomi focuses the intelligent experience on the device that users most often interact with - the mobile phone. HyperAI, image enhancement, AI translation, and cross-device collaboration form the basic form of "intelligence in the palm". Taking the Xiaomi 17 Ultra as an example, its 1-inch main camera combined with LOFIC HDR technology and AI scene analysis maintains excellent details and dynamic range even in complex lighting conditions. At the same time, HyperAI provides capabilities such as image enhancement, AI writing, voice recognition, and dialogue translation, while HyperConnect enables instant sharing of photos, videos, and mobile phone screens between different devices, allowing the intelligent experience to continue naturally across multiple terminals.

The intelligence continues to extend from handheld devices to the AI Tomorrow section. The travel scenario demonstrated by the SU7 Ultra allows AI to participate in driving, perception, and spatial linkage. Alongside it, the Xiaomi Vision Gran Turismo concept electric supercar, which made its physical world debut at MWC, was also on display. This future vehicle designed for the "Gran Turismo" series presents a design language of "shaped by the wind" with its drop-shaped cockpit and sculptural aerodynamic structure, making Xiaomi's imagination of future travel forms concrete on the booth.

Different from single AI hardware, Xiaomi's terminal system itself covers personal devices, mobile spaces, and living spaces, and intelligence can extend along this trajectory.

Scale not only brings a user base but also real data and usage feedback. New capabilities can be quickly implemented, and the feedback in turn promotes model optimization, forming a self-reinforcing cycle.

When the model and hardware are combined, Xiaomi has a capability that other AI manufacturers don't have - the ecosystem.

The "full ecosystem of people, cars, and homes" itself fits the real life routine of users. Users plan their trips on their mobile phones, continue to use the same account and voice assistant in the car, and when they return home, the devices automatically switch scenarios.

In this context, the whole-home intelligent system Miloco, driven by a large model and first shown overseas at MWC 2026, has become one of the specific manifestations of this strategy.

Based on the MiMo model capabilities and the large scale of IoT devices, Miloco enables Xiaomi's intelligent capabilities to enter home spaces and participate in environmental understanding and device linkage. Users can express their needs through dialogue without manually setting complex rules. After the camera recognizes changes in behavior, the system automatically triggers lights, air conditioners, or cleaning devices. Multiple devices form a linkage according to the scenario and automatically adjust their states.

From the model base to the terminal scale and then to the coordination of life scenarios, Xiaomi presents a trinity form. Different from companies that only provide an AI device or a chat interface, Xiaomi is more like building an intelligent path through life, allowing model capabilities to truly enter mobile phones, cars, and home spaces. The "full ecosystem of people, cars, and homes" has landed from human imagination and continues to operate in the real world.

Xiaomi's ecosystem: a "Chinese path" for AI

As the industry enters the application period, the differences among AI manufacturers are becoming clearer.

Alibaba is more rooted in business and enterprise scenarios, extending large model capabilities around e-commerce and cloud infrastructure. ByteDance relies on its content distribution and creation ecosystem to let AI improve production efficiency and content generation capabilities. Tencent embeds intelligence into communication and office processes around its social and collaboration systems. Most of these paths occur in the digital world - revolving around information, traffic, and service circulation.

In contrast, Xiaomi's exploration is closer to the physical space. Mobile phones, cars, and home devices are not the "super assistant entrances" that people pursued in the past but real carriers that users touch, drive, and live in every day.

This choice not only responds to users' real and continuous life needs but also is based on Xiaomi's years of ecosystem accumulation.

Over the past decade or more, Xiaomi has been deploying around hardware and the ecosystem. Starting from smartphones, building an IoT system covering more than 200 categories, and then entering the field of intelligent cars, Xiaomi's core ability has always been to implement complex technologies on a large scale and replicate them in the global market. Behind the figures of 740 million monthly active mobile phone users and over 1 billion connected IoT devices is Xiaomi's terminal network that has been operating stably for many years and real scenarios with continuous feedback.

Since 2024, Xiaomi has focused on investing in large models as a new underlying capability. The MiMo series of models have been comprehensively developed in the directions of language, multimodality, voice, and embodiment and have entered the first echelon in the open-source system. At the same time, the underlying capabilities such as chips, operating systems, and robots continue to be advanced.

After the AI competition enters the application period, the significance of this layout is magnified.

The technology exploration period is gradually coming to an end, and parameter scale is no longer the only measure. All AI manufacturers must think about how to stably and continuously embed model capabilities into the real world, how to run on real terminals, and how to coordinate in the manufacturing process, travel scenarios, and home spaces.

These are exactly the areas that Xiaomi has been good at for a long time.

Globally, at the critical point from the first half to the second half, there are not many manufacturers that have both self-developed model capabilities, edge-side hardware scale, and cross-scenario ecosystem integration capabilities. Xiaomi's path is based on long-term hardware accumulation and global channel expansion, and with the continuous investment in AI in recent years, it has a unique position in the application level of the physical space.

The investment also further reflects the determination. In 2025, Xiaomi invested about 7.5 billion yuan in the AI field, accounting for a quarter of the annual R & D budget. It plans to invest a total of 200 billion yuan in R & D in the next five years. The MiMo series of models continue to be open-sourced, and the exploration of architectures such as HySparse tries to find a balance between efficiency and cost. "Xiaomi has made large investments in the most fundamental aspects, whether it's chips, AI, operating systems, or robots. These technologies have strong commonalities," Lu Weibing said in an interview.

Whether it's the past accumulation, future investment, or current ambition, Xiaomi's actions all point to an answer to a question: After AI becomes a basic capability, can Chinese technology companies form new competitive advantages in the application level of the real world?

In the past, Xiaomi's advantages came from large-scale manufacturing and global channel capabilities. Today, this accumulation begins to be combined with self-developed model capabilities. In the future, this combination may have an impact in more life scenarios.

If the keyword in the first half is catching up and breakthrough, then the keyword in the second half is closer to integration and implementation. In this change, Xiaomi's path is not only about the strategic choice of an enterprise but also provides a sample for the position of Chinese technology companies in the global AI competition.

There is past accumulation, present layout, and future imagination space. This path is still unfolding.