Cheetah Mobile (CMCM) Q1 2026 earnings call transcript

Thank you, Albreeder. Welcome to Cheetah Mobile's first quarter 2026 earnings conference call. We said today our company's chairman and CEO, Mr. Fu Shen, and our company's director and CEO, Mr. Thomas Jennings, following management's prepared remarks, will come back to the Cheetah section. Please note that the management's craze will be presented by an AI agent. Before we begin, I refer you to the same however statements in our earnings release, which also apply to our company's call today, as we will make the word we keep saying.

2026 remains an important transition year for Cheetah Mobile. We are continuing to evolve from a traditional internet company into a company focused on AI enabled applications or AI agents and robotics. More importantly, we believe we are gradually moving from capability building into early stage commercial validation. Our focus is not only on developing the high capabilities, but on turning these capabilities into practical products for real business scenarios, helping customers deliver better ROI. Starting from this quarter, we are separating our robotics and others' business into an independent reportable segment. In the first quarter, revenue from robotics and other business increased 176% year-over-year to 51 million RMB, approaching 20% of total revenue. At the same time, our just about putting us from this segment narrowed by 57% year-over-year.

Customer demands remain strong, and we expect robotics and others with India to grow strongly in 2026. In Q2, our robotics and other revenue will continue growing on both year-over-year and culture-over-culture basis. Today, our robotics business mainly focuses on commercial scenarios with real customer demand and clear long-term value, including reception, data tours, and intelligent service applications. Our smart personal mobility is another important step for us. This product extends our robotics and AI capabilities into personal mobility and healthcare-related scenarios. More importantly, it further validates that our robotics platform can expand beyond commercial service robots into broader consumer applications. We are encouraged to see recognition from leading industry partners. During the second quarter, we started initial product shipments to a top local designer and manufacturer of mobility products, as well as a leading airline mobility scooter manufacturer in China.

We are seeing encouraging early market feedback and initial commercial traction. Moving to our agents, we're seeing strong customer adoption. We work closely with Google Cloud and AWS, helping enterprises serving international markets access AI models and use multi-cloud environments more efficiently. In one, two, two, six, revenue from our cloud and AI infrastructure services as a part of global enterprise service revenue increased 68% year over year, contributing 18% of total revenue. Daily energy token usage has increased more than 20 times since January of 2026, exceeding $400 billion in May. We expect this revenue growth to continue. We also kept building EZQL. So early, we believe it will help customers deploy AI apps and boost productivity. The two flex-growing businesses, namely Robotics and others, as well as Cloud and AI infrastructure, already accounted for 38% of our first quarter revenue, and we expect their revenue growth and revenue contributions to continue growing in the coming quarter and to exceed more than 50% of our total revenue in the second half of this year.

During the quarter, revenue from our advertising agency business within the global enterprise services segment was affected by policy changes from certain overseas advertising platforms. We believe this revenue decline was primarily driven by external factors rather than changes in customer demand. This was the primary reason for the company's widening year-over-year operating loss in the first quarter. Our Internet services business continues to provide important profit and cash flow support for the company. In the first quarter of 2026, our Internet service business generated approximately 15 million RMB in adjusted operating profits in 2026.

All financial figures presented in RMB. During the first quarter of 2026, we continued focusing on operating discipline, improving revenue quality, and maintaining financial flexibility as we invest in AI and robotics initiatives. Total revenue remained relatively stable year-over-year at $259 million RMV during the quarter, while Internet service revenue declined due to continued weakness in online advertising.

Quality of our revenue must continue to improve. Within the Internet service segment, revenue from Internet value added services continued to grow steadily at 8.2% year-over-year, contributing 72.8% of segment revenue. During the larger portion of the internet value added services, our internet service revenue is becoming increasingly predictable. More importantly, the internet service business remains profitable and continues generating stable cash, which provides an important financial foundation for our long-term AI and robotics investments.

Turning to our robotics and other segment, starting from this corner, We begin recording the robotics and others' business as a separate segment to present the operating progress of this business. Historical results previously reported on the AI and others are now presented as robotics and others, as well as global enterprise services. During the first quarter, revenue from robotics and others increased significantly year-over-year, with revenue increasing 175.9% year-over-year to 51.2 million RMD, accounting for 19.8% of total revenue. Adjusted operating loss from this segment narrowed by 57.1% year over year, reflecting continued improvement in operating efficiency and commercial execution. According to global enterprise services, this business remains strategically important to the company.

In addition to profitability attribution, it provides us with valuable enterprise customer relationships, overseas operating experience, and real-world deployment scenarios for AI-related services. During the quarter, revenue from the advertising agency business was affected by policy changes from overseas advertising platforms, which impacted year-over-year segment revenue performance. However revenue from our growth and AI infrastructure services business increased by 68 supported by increasing advertising demand for AI cloud and its open management services Coming to profitability operation loss was on the million during the quarter Compared with $26.5 million RMB in the same period last year.

Increased mainly reflected lower profitability on the internet and global enterprise services business. Following rapid declines in online advertising and advertising agency services, as well as our continued investment in AI and robotics initiatives. More importantly, most of the internet service and global enterprise services business remain profitable due to the quarter. The Internet service business generated approximately $15.2 million RMD in adjusted operating profit, while global enterprise services generated approximately $13.8 million RMD in adjusted operating profit. We also maintained a strong balance sheet. As of March 31, 2026, we had approximately $186 million United States dollars in cash and cash equivalent, as well as over $100 million United States dollars in long-term investments. We believe our financial position provides sufficient flexibility to continue investing in AI and robotics with a discipline and sustainable approach.

Looking ahead, our financial priorities remain consistent. A. Maintaining operating discipline. B. Improving revenue quality on operating efficiency. C. Supporting long-term investments by preserving financial flexibility. Overall, we believe the company continues moving toward a more sustainable and balanced operating structure as our AI and robotics business is gradually scale. Thank you.

Okay, let me answer. Thanks, Thomas, for your question. I think you also pointed out a very important issue in the robotics industry, which is the issue of insufficient training data today. The rapid development of AI has given us very high expectations for the robotics industry, believing that today's AI capabilities have improved, and robots should soon be able to achieve various behavioral capabilities. But in fact, I don't think so, because the development of AI agents, including the development of large language models, is actually built on the development of the Internet for two or three decades. The Internet essentially forms the basic training data of large language models. It is a very high-quality data set, and the biggest problem in the robotics industry today is the lack of data.

And many ways are being tried today, with many manufacturers trying to use training data. including data migration, simulation training, and so on. However, there is a very serious problem. The physical world is much more complex than the laboratory environment and the simulator environment. So today, whether it's data migration, collection, or truly migrating to different ontologies, its adaptability will be a huge challenge. Let me give you an example. Today's I think Tesla's FSD is already very good. But in fact, some older versions of Tesla's own car cannot install the latest FSD, so indeed data is a very big problem.

I also very much agree with what you said. The data continuously generated in the real deployment environment is actually very important for the robotics industry from our own experience. Let me give you two examples. One aspect is our voice interaction capability in different environments, which is actually closely related to our long-term exploration in various scenarios. Different noises, different environments, multiple people, and so on. We have made some optimizations and training on the data. Therefore, the interaction effects of our interaction robots, including reception, are leading in the industry today.

We have a reputation of our own in the industry. Another example is the mechanical mobility. A very simple robot can navigate indoors from point A to point B. It is similar to a small low-speed driverless vehicle. How to use cheat chips and sensors to achieve automatic obstacle avoidance in different environments. In fact, all of these can only be achieved based on massive amounts of data. We recently launched a smart wheelchair, which we just mentioned. We started mass production in May, and now it seems that in overseas markets, especially in Europe, the sales momentum is quite good. In fact, for a traditional wheelchair product like this, to achieve obstacle avoidance and assist driving, many manufacturers, including some startup manufacturers, want to achieve this kind of assisted driving capability.

But to create a prototype and truly achieve good passing ability in many environments, it actually requires quite a lot of effort. This is related to the fact that we have deployed many robots in many environments over the years, regardless of the surface conditions, such as carpets or floors. We have also enhanced the reflection of walls, all of which have accumulated over time. There is also continuous algorithm optimization based on actual scenarios. Therefore, our wheelchair can truly achieve lower cost, highly assistive driving capabilities. It has also received... So, at this stage, the value chain is definitely in this regard. But I want to say the first is why I think it is not the model layer, because although the model, there is very fierce competition.

But what we see now is that the gap between models is not too wide, and it is not easy to widen. Today, for example, the models of China and the United States, we think there is probably a gap of about half a year. And this gap is probably such a process. And there is no side that pulls the other side away. And among large manufacturers, I think the gap is a bit like even flow.

Of course, today's models are also in the early stage. And in the future, I think with the continuous increase in production of inference chips and training chips, the training costs will gradually decrease. So I think the model layer will be an infrastructure, but in the long run, it will not be monopolized. And with the continuous improvement of the model's capabilities, now we can see that many models, even if they are not top models, Specialized models have been adapted to some daily tasks and actually achieved very good results. For example, some open source models in China this year have seen a significant increase in the amount of calls, and I think the core reason is that they offer great cost effectiveness.

They have achieved high completion rates in some tasks. Therefore, I even think that in the future, various specialized models will continue to emerge. Of course, this will take some time. The second infrastructure layer we do not fully participate in but we also see that because we have our own cloud business and we have token clients consuming here the growth is also very fast Still I think this is a state of mismatch between supply and demand at this stage but eventually the infrastructure will also enter an economy of scale And for applications, today AI can actually reshape almost all applications. So there are huge opportunities in the application layer today. Whether it is the industry we are doing like robots, we have been doing it for a long time, but we are still very firmly optimistic.

When the capabilities of the model continue to improve and the application of robots is wider, there are many sayings that it may be a bigger industry than the automotive industry. There are also many opportunities at the software level, which I will not expand on here. Even today, when we look at some large model companies, their valuations are very high or excellent. In fact, they have truly delved deep into a certain application, such as the programming of disabled diffusion, and the rise of Quad is actually an application. Its application is a coding application. It has made the coding application good enough, rather than just providing an API for you to call, but its agent has been well-developed, including OpenCloud that emerged at the beginning of this year.

We have also developed products like EasyCloud. I think there is still a large room and opportunities in the application layer. Well, thank you.

From my understanding of the robotics industry today, I believe that in the short term or within the next three to five years, the possibility of a particularly versatile robot appearing is very, very low. This is limited by both the so-called model capabilities and the entire hardware industry chain. The update of the hardware industry chain is actually relatively slow, and it involves some of the most basic physics and materials, as well as the underlying logic of physical laws and materials. So I believe today that the core skill barriers in the integrated industry in the future still lie in the true scenario operation capabilities. And in terms of client network, if today we can have enough scenarios and have a good client network so that our products can really be used in these scenarios, we can accumulate our own unique experience or data.

The first question has been answered, which is that we can optimize it. And this optimization enables the product to provide better cost-effectiveness to truly meet users' needs. The machinery industry is very hot, but when it comes to business implementation, clients don't care whether you are a robot, a machine, or a human. What they care more about is cost-effectiveness, ROI, input, and output. This has been very significantly reflected in our operations in recent years. So whether it is in the media, you see a lot of amazing things before, but you will find that it is really in actual scenarios very few, without going through actual scenarios. Let me reiterate this. The operation of robots in the physical environment, whether it is actions or work, its complexity, is actually much higher than that of autonomous driving of cars.

So in this case of very high complexity, I think in practical application scenarios today, in the operation of scenarios and customer networks, a vertical and penetrating point can be formed. It is much more important than a generalized machining model, because today I don't think Generalized models and machines can quickly complete the ROI required in these vertical scenarios. Okay, thank you.

I think this is a very broad question. I think the ultimate mode of enterprise-level AI products should come from a deep understanding of user needs and a deep understanding of the industry, and then form an extremely high level organizational capability. Because the points you mentioned today are also realistic. In terms of the capabilities of the model itself, it seems that one thing rises and the other falls. Then, cost effectiveness is also increasingly being brought up. So what is the essence today? It actually allows enterprises to save a lot of money that used to be spent on non-commercial insights, user insights, and truly focus on understanding user needs. So the real load comes from your keen insights into user needs and quickly launching your products and services and improving your products and services.

So we often talk about the AIAT5 organization. In essence, it is to use AI to reconstruct the internal organizational processes of the enterprise and to quickly and efficiently achieve the operation of the enterprise and to launch its own products and services more efficiently and quickly. For example, if you pay attention, we have launched various product services in the past year, much more than in the past. But our investment in R&D has decreased a lot from the perspective of cost. Although there is still room for improvement, this is an example. So when you launch products and services so quickly, where is your real deal? It comes from users' demands. You can really find users' demands and quickly launch and quickly respond to users' demands.

By the way, we have also launched some corresponding services and courses for the organizational construction of AI technologies for the enterprise version, and shared some of our experiences with our clients. Now, some big clients have started to sign contracts. Operations have also begun. The essence of business competition lies in efficiency and insight into user demand. And I believe AI products can accelerate the arrival of these two points. Okay, thank you.

I think the core reason why there is still such a large amount of customization and manual services today is that AI is still in its early stage, although we all see in moments of various media that most people's understanding of AI and its use are still insufficient. I think only a few people today can really make good use of AI. So this is a generation gap. Today's In a historical enterprise we need to do customized development and manual services for this district Traditional SaaS has been developing for many years and it has condensed many things in the code so it seems that in many cases it belongs to standardized delivery I think as everyone gradually understands AI, the entire staff are getting more and more proficient in AI applications.

The proportion of this service model will continue to decline. Our company has already achieved a model where all employees are using AI to write code, and some of our internal systems are also starting to use AI to be written directly by the business department, rather than relying on CAD software and the service department. So the most critical change in this process is, on the one hand, I think the model capabilities are constantly increasing.

And today, for example, a very important feeling we have this year is that today the business department is writing some internal software and services. And when using the model, we feel that the model capabilities have been improved a lot compared to last year. And many of them may have been more of a demo before, or a demo-level product that can already be used internally. The model capabilities will continue to increase. Another thing is that our organizational structure today is still based on the traditional one, based on industrial software. I think with the continuous emergence of emerging companies, new AI-native organizations are emerging, and the traditional standardized SaaS model will be broken. So what we provide to our customers today is no longer the traditional type of service, but more of training, training for our clients, employees, and assessment of AI capabilities to help them transform their AI organization.

I think this change is the most critical, Which means that companies need to change their organizational structure and demands for employees based on AI. Thank you.

I believe I have made my views on humanoid robots quite clear in the media. I think humanoid robots will not be able to replace humans in commercial applications beyond performances in the next three to five years, We have a lot of work in the past five years, no matter in factories or service industries, or even in households. The difficulty of developing humanoid robots is extremely high. We have wheeled robots in robotic arms, like XM3 factory. These robotic arm products have been steadily growing in recent years, and has shown good growth in Q1 this year. The real robots are also doing well with practicality, cost-effectiveness, and indoor navigation technology all reaching a mature stage. Therefore, I believe they will also see rapid growth.

This has been my cue. I believe robots should evolve from specialized vertical robots that continuously grow and gather data until they are advanced enough, and then they will undergrade.

Congratulations to the more general form. As for bipedal robots, I don't think they are needed in most scenarios. There's no need to add such cost and complexity, including its reliability. So this is my opinion, and we have reiterated it many times, that what we care most about in making robots is the commercial lending that can really be accepted by the market and is really paid by the market entity, not just some project lending or some integrated project. So I think the real robots will gradually be matched with Product B in the future, and for a long time, it will be the main form of humanoid robot development.

Thank you.

Actually, home robots are a broad concept. If you really talk about home robots, the only breakthrough is a sweeping robot. It's also called a robot, right? But if you consider it a robot that can do more half-jew staff than adults, I think the first reason why we make intelligent wheelchairs is that, in our view, intelligent wheelchairs are robots. But previously, our robots were used for delivery, and intelligent wheelchairs can also be understood as delivering people. So I think the first step of the application is mobility, the ability to move from A to B. The second is to add some functions on this mobility, such as adding the ability to sweep the floor for a sweeping robot.

What we see now is companionship. Helping you achieve some family control, controlling it with voice and integrating it into robots, helping you make some plans, and being a good companion. These are all part of the same direction. Actually, it can also be said that our wheelchair products have such functions, including the companionship function for the elderly, which will soon be launched on our HPP. I think if it's like what everyone imagined, such as the ability to do housework, I don't think it's possible to achieve it within two to three years. Because we have our own robotic arm company. And our robotic arms are used in many scenarios, whether in industrial or commercial settings. And even in scenarios like commercial dishwashing, we have seen such cases.

Today, it's important to note that interacting with the physical world is extremely complex for robots, not because they can perform certain actions, but because of the stability that follows and the success rate.

Even the success rate of picking up a cup today is not 100% for any company, even in a kitchen environment, for home use. If the success rate is 99%, you'll still accumulate broken cups. This negative impact is quite significant, not to mention if it enters a household. There will be issues like falling, bumping into things, or hitting people. There's also the reliability of its quality. We expect the home appliance to work fine for several years after purchase, but for a complex robot, ensuring quality over a long period of time without malfunctioning is a tough challenge for many robotics companies today. Therefore, I believe that when it comes to home robots, we should be more pragmatic.

Our view is that robots should be able to truly provide companionship for the family and assist the elderly and people with disabilities in moving around. I think this is a great, great two-direction. Thank you. Okay, Operator, please check if there are any further questions. And if not, we can conclude the meeting.