Cheetah Mobile's Q2 2025: Contradictions Emerge on AI Commercialization, Costs, and Robotics Profitability

The above is the analysis of the conflicting points in this earnings call

Date of Call: None provided

Financials Results

• Revenue: RMB 295 million, up 58% YOY and 14% QOQ
• Gross Margin: 76%, compared to 65% in the prior year and 73% in the prior quarter
• Operating Margin: Operating loss of RMB 11 million, improved 86% YOY and 58% QOQ; non-GAAP operating loss RMB 2 million, down 97% YOY and 86% QOQ

Guidance:

• Expect to maintain fast growth in H2 2025; AI & other segments targeted at ~100% YOY revenue growth; internet business stable/profitable.
• Confident on a near-term return to profitability; internal goal to achieve Q3 overall profitability, though no formal guidance provided.
• UFactory acquisition to add incremental revenue from H2 2025; plan to scale globally via 100+ partners.
• Robotics commercialization to advance steadily; no mass deployment in coming quarters; focus on ROI-positive, scalable use cases.
• Continue disciplined investment in AI tools and robotics; strong cash and zero debt provide flexibility.

Business Commentary:

* Revenue Growth and Profitability: - Cheetah Mobile reported a revenue of RMB 295 million in Q2 2024, marking a 58% year-over-year increase and 14% quarter-over-quarter growth. - The growth was driven by a 39% year-over-year increase in the internet business and an 86% year-over-year increase in AI and other segments.

• AI and Technology Integration:
• The company's AI and other segments, led by the AI tool DeepFlow, showed encouraging early user adoption, validating product-market fit.

• This was supported by the integration of AI agents into existing apps like Dubai Antivirus and PDF tools, enhancing their functionality.

• Service Robotics and Acquisitions:

• Revenue from service robots continued to contribute to growth in the AI and other segments.

• The acquisition of UFactory One, a profitable robotic arm company, is expected to enhance Cheetah Mobile's distribution network and global partnerships.

• Internet Business Stability:

• The internet business generated RMB 4,700 million in operating profit for the first half of 2025, with an adjusted operating margin of 14%.

• This stability is attributed to the shift from an ad-centric model to a user-subscription-driven model, improving user engagement and retention.

• R&D Efficiency and Strategic Focus:

• R&D expenses accounted for 24% of AI and other segments' revenue in Q2, down from 39% in the previous year.
• This reduction is due to the company's focus on ROI evaluation and efficient use of AI tools, reducing compute-intensive directions and streamlining R&D processes.

Sentiment Analysis:

• “Revenue grew 58% year-over-year… Gross margin improved to 76%… Non-GAAP operating loss declined to RMB 2 million… We are on track to reach profitability in the near term.” Management highlighted a stable, profitable internet business, ~100% YOY growth targeted in AI/other for H2, and confidence in Q3 performance, while noting robotics mass deployment will take time.

Q&A:

• Question from Thomas Zhang (Jefferies): How sustainable is the ~40% internet business growth into H2, what are revenue/profit drivers, and what is the normalized revenue/margin range excluding AI tools?
  Response: Growth is sustainable via subscription-led model, channel expansion, and AI features; internal goal is H2 revenue growth and margin above last year, with margins fluctuating modestly but generally improving; internet remains a cash generator.

• Question from Becky Wei (Citi): Please detail robot sales performance, split between voice robots and robotic arms (UFactory), and whether high growth can persist in H2.
  Response: Voice-robot growth is driven by LLM-powered interactions improving UX; expect continued demand; UFactory is profitable with strong overseas exposure and will scale via Cheetah’s network; robotics remains early, with disciplined, ROI-led expansion.

• Question from Nancy Liu (JP Morgan): With Q2 near break-even, can you achieve overall profitability in Q3, and what’s the biggest obstacle to robot business breakeven?
  Response: Management is confident on Q3; main obstacle is identifying scalable, repeatable, ROI-positive use cases; focus on strategic prioritization, strict project-level ROI, and leveraging UFactory’s proven niches.

• Question from Brenda Zhang (CICC): After integrating AI agents, are there quantifiable changes in user behavior like time spent, retention, or willingness to pay?
  Response: Robots show clearly higher satisfaction and repurchase; in internet apps, early tests indicate better engagement and conversion, but data is early; broader rollouts planned over the next two quarters.

• Question from GG Zhang (GF Securities): What is the long-term moat for AI tools amid intensifying competition?
  Response: B2B moat comes from embedding in workflows and leveraging proprietary data plus sales reach; B2C moat is dynamic—speed of iteration and user mindshare, not permanent barriers.

• Question from Zhang Group Lee (Guotai Junan Securities): Progress and timeline for wheeled robots with arms, and other new product forms and use cases?
  Response: Monitoring demand-led scenarios with no fixed production timeline; technical capabilities exist in bases and arms; will launch when ROI and scalability are proven; exploring larger validated use cases.

• Question from Zhang Huang (Everbright Securities): Can you quantify R&D efficiency gains from AI tools and discuss future efficiency focus?
  Response: AI-enabled small teams markedly boosted output (e.g., 2 people delivered work previously needing 5–10); aim to build an AI-native organization; optimize efficiency without simply cutting R&D.

• Question from Guang Peng Zhang (Sealian Security): How does your robotics strategy and business model differ from peers?
  Response: Prioritize ROI and repeatable deployments over frontier tech for its own sake; core strengths are software UX and voice agents; augment hardware via acquisitions and fit solutions to scalable scenarios.

• Question from Yan Peng Zhao (Haitong Securities): Will you pursue more M&A given sizable long-term investments?
  Response: Open to deals with strong strategic/cultural fit like UFactory; evaluate rigorously on synergy, strategy, culture, and valuation; portfolio provides optionality for targeted capability building.

• Question from Johanna Ma (CMBI): How do AI tools’ cost structures differ from traditional tools, and what is the ROI outlook?
  Response: AI-assisted coding slashes dev cost/time; model token costs are higher but declining rapidly; users show strong willingness to pay; expect sustainable unit economics as pricing models mature.

• Question from Jack Yang (Mizuho): Is a user-paid model for AI tools viable in China and what’s the timeline?
  Response: Yes—domestic willingness to pay has risen; subscriptions already ~60% of internet revenue; focus on delivering clear value; pursuing both domestic and overseas monetization.