HK Stock Market Barometer | HK stocks continue to fluctuate and pull back! How much room for recover
Kuaishou Q2 2026 Earnings Conference Call
[AI Key Takeaways]
Financial Performance
- Total revenue in Q2 2026 reached RMB 35.5 billion, a year-on-year increase of 4.1%
- Adjusted net profit was RMB 3.9 billion, with an adjusted net profit margin of 11%
- Online marketing services revenue amounted to RMB 20.6 billion, up 4.4% year-on-year
- Other service revenue reached RMB 6.2 billion, a year-on-year increase of 18.5%
Business Progress
- Kuaishou app averaged 412 million daily active users and 797 million monthly active users
- AI-related operating revenue is estimated to exceed RMB 850 million, representing a year-on-year growth of over 200%
- Launched the industry's first video model supporting native 4K direct output
- Employee adoption rate of AI products exceeded 92%, with AI-generated code contributing 60% of R&D staff's output
Next Quarter Guidance
- The second half is expected to face short-term pain points, with revenue under pressure and continued AI investment impacting profit margins
- E-commerce marketing service revenue and commission income are expected to face year-on-year pressure in the second half
- Targeting positive group-level free cash flow in the second half
- Total shareholder returns for the full year 2026 are projected to surpass last year's level
opportunity
- AI technology continues to reduce content production costs, driving rapid expansion in the supply of content such as short dramas
- Sub-sectors such as lifestyle services still have room for growth, while the commercialization potential of gaming content remains to be unlocked.
- AI is being comprehensively applied across the entire marketing service workflow, enhancing client ad delivery efficiency and ROI.
- The AI business may secure independent financing, thereby boosting its competitiveness within the industry.
Risks
- The macroeconomic environment is complex, with limited visibility into external conditions.
- Industry competition is intense, with sectors such as AI applications facing significant pressure from high base effects.
[AI Conference Transcript]
Operator
Ladies and gentlemen, good day, and welcome to Kuaishou Technology's conference call for the second quarter and interim results of 2026. Please note that simultaneous English interpretation will be provided for the management's remarks. You will need to access the English channel to listen to the interpretation by third-party interpreters. Please be advised that today's conference is being recorded. I would now like to invite Mr. Matthew Zhao, Vice President of Capital Markets and Investor Relations, to speak.
Matthew Zhao
Thank you, and hello everyone. Welcome to Kuaishou Technology's conference call for the second quarter and interim results of 2026. Joining us on the call today are Mr. Cheng Yixiao, Co-founder, Chairman, and Chief Executive Officer of Kuaishou Technology, and Mr. Jin Bing, Chief Financial Officer of Kuaishou Technology.
Before we begin, we would like to remind you that today's discussion may contain forward-looking statements, which involve numerous risks and uncertainties. Actual outcomes and results may differ materially from those discussed today. Except as required by law, the Company assumes no obligation to update this forward-looking information.
For all important information regarding this conference, including forward-looking statements, please refer to the Company's public filings or the announcement of the second quarter results ended June 30, 2026, released on the Company's website earlier today.
During today's conference call, management will also discuss certain non-IFRS financial measures. These are provided for supplementary consideration only and should not be regarded as substitutes for the company's financial performance metrics prepared in accordance with International Financial Reporting Standards (IFRS). For definitions of these non-IFRS financial measures, their reconciliation to IFRS financial performance, and related risk factors, please refer to the company's Q2 2026 earnings announcement.
Management will primarily communicate in Chinese during today's conference call. For the first part, consisting of management remarks, third-party interpreters will provide simultaneous English interpretation. For the second part, the Q&A session, third-party interpreters will provide consecutive English interpretation. The translation is intended solely to enhance meeting efficiency; in case of any discrepancy between the translation and the original Chinese content, the original Chinese statements by management shall prevail.
Finally, unless otherwise specified, all monetary units mentioned in the conference call are in Renminbi (RMB). Now, let us invite Mr. Cheng Yixiao to speak.
Cheng Yixiao
Hello everyone, and welcome to Kuaishou's Q2 2026 earnings conference call. In Q2 2026, amidst a complex macroeconomic environment and industry competition, we adhered to long-termism, remained committed to our AI strategic investments, and achieved high-quality growth.
In Q2 2026, Kuaishou App's average daily active users (DAUs) reached 412 million, and total revenue increased by 4.1% year-over-year to RMB 35.5 billion. Core commercial revenue, which includes online marketing services and other services primarily driven by e-commerce and potentially AI, grew by 7.4% year-over-year.
Adjusted net profit was RMB 3.9 billion, with an adjusted net profit margin of 11%. The company's overall profitability remained relatively stable, further highlighting its operational resilience. Next, I will provide a detailed overview of the progress in key business segments for Q2 2026.
First, regarding the company's AI strategy and progress in large models. In Q2 2026, Keke AI continued to pursue its vision of enabling everyone to tell compelling stories with AI. Through cutting-edge model capabilities, breakthroughs in professional product feature upgrades, and the expansion of an international creative ecosystem, it has further consolidated its global leadership in multimodal video generation large models.
At the model and product level, the Keke AI model introduced native 4K direct output functionality, marking the industry's first video model to support native 4K direct generation, allowing for one-click creation of high-definition, high-texture videos. This feature targets professional video industries such as film and advertising, enabling clients to directly output high-resolution visuals in real-time without complex post-processing, thereby achieving cinema-industry-level visual effects.
Meanwhile, Keke AI launched its Model 3.0, which significantly boosts creation efficiency and reduces costs while ensuring stable dynamic quality output and precise audio-visual synchronization. The capability of agents to orchestrate Keke AI for batch content creation has further expanded the application boundaries of Keke AI in workflow automation and intelligent scheduling scenarios.
Keneng AI’s continuous integration of video creation capabilities empowers professional content production, with its technical prowess and creative outputs widely recognized within the industry. At the 2026 Cannes Lions International Festival of Creativity, advertisements generated by Keneng AI won one silver and two bronze awards, demonstrating its creative capabilities and earning recognition from top-tier creative evaluation systems.
At the 2026 Beijing International Film Festival, multiple works created using Keneng AI, including 'Paper Airplane,' were shortlisted in the AI Execution category. 'Shenbi' (The Magic Brush), with renowned sci-fi author Liu Cixin serving as literary supervisor, was named a Work of Annual Interest in the Short Drama and Micro-Short Drama category of the Short Video Unit, fully showcasing Keneng AI’s strength in empowering professional film and television production.
Driven by breakthroughs in model capabilities, improvements in product features, and the deepening of application scenarios, Keneng AI’s commercialization has maintained a trend of rapid growth. In the second quarter of 2026, Keneng AI’s revenue exceeded RMB 850 million, representing a year-on-year increase of over 200%, continuing to lead the global commercialization process of the AI video production industry.
In the second quarter of 2026, we continued to make significant progress in the research and application of general-purpose large models. We released the new version of our multimodal large model, Keneng VR 2.0, which successfully unlocked deep perception for ultra-long contexts of up to 256K tokens, achieving nearly lossless inference capabilities in temporal perception for long-form videos.
Keneng AI has also introduced agent collaboration mechanisms, possessing the potential to execute tasks such as code parsing and tool adjustment. We launched Agent X, an autonomous agent proxy, for industrial recommendation systems, enabling them to autonomously complete recommendation model and strategy design, performance evaluation, and experience accumulation, thereby significantly improving the iteration efficiency of recommendation algorithms.
In the second quarter of 2026, regarding the application of large model technologies, we built scenario-based capabilities for producing and configuring marketing materials in online marketing service scenarios. By providing more suitable material production capabilities tailored to the needs of different industries and clients, we drove a year-on-year increase of over 70% in AIGC short-video marketing consumption.
We expanded the application of generative recommendation large models and intelligent bidding large model technologies to multi-scenario environments such as live streaming search, enhancing the effectiveness of marketing material recommendations and leveraging client marketing budgets. In terms of organizational efficiency and empowerment in the second quarter of 2026, Kuaishou’s self-developed general-purpose agent product completed key integrations with internal systems via skill modules and began serving all employees comprehensively.
In June, the proportion of Kuaishou employees using various self-developed AI agent products exceeded 92%, while the AI code contribution rate among R&D technical staff reached 60%. Meanwhile, Kuaishou Wanqing, as an enterprise-level large model service and development platform, integrates high-performance model inference, low-cost model customization, and fully managed services. In addition to supporting Kuaishou’s internal AI operations, it provides large model infrastructure services to external clients.
On Open Writer, a global large model aggregation platform, Kuaishou Wanqing offers several open-source models that rank among the top in terms of call volume.
Secondly, regarding the user and content ecosystem, the average daily active users (DAU) and average monthly active users (MAU) of the Kuaishou app reached 412 million and 797 million, respectively, in the second quarter of 2026. We improved customer acquisition efficiency for new users through AI-driven intelligent delivery methods and enhanced the overall retention rates of both new and returning users.
By iterating our traffic distribution system, we have better safeguarded the experience of highly active core users. We continue to optimize social features, with the number of users engaging in two-way private messaging increasing by over 15% year-on-year. Meanwhile, we have focused on improving Kuaishou's foundational functionalities, achieving systematic optimizations in video playback smoothness and intelligent interaction, thereby comprehensively enhancing the user experience.
We remain committed to leveraging the power of our community, deepening our engagement, and highlighting the differentiated, high-quality content that defines the core of the Kuaishou community. In June and July of this year, capitalizing on the social buzz surrounding the World Cup, we launched our native IP, the Kuaishou Sports God Cup. In addition to curating hot movie and TV content, we introduced several original campaigns, such as the Xiangyashan Football Tournament and the Dream Chasing Youth Football Match.
Through trending topic operations, fun interactive activities, and mass co-creation, we have created a new sports arena where all users can participate and express themselves. Content related to these events achieved a total of 68.2 billion impressions on the Kuaishou app, with cumulative live stream viewership reaching nearly 360 million.
Furthermore, we have deepened innovative cooperation mechanisms for copyrighted content, delivering more high-value content consumption opportunities to users through copyright operations. We introduced the broadcast rights for the 2026 CBA season using an e-commerce livestreaming model and extended the paid livestreaming model to online music performances. In April, we launched the TFBoys East Road Boys Group concert, achieving single-event sales exceeding RMB 10 million, thereby realizing a win-win scenario for both content and commerce.
Additionally, the paid livestreaming model has stimulated community-organized sports events, particularly creating regional scale effects in Northwest China.
Thirdly, in terms of online marketing services, revenue reached RMB 20.6 billion in the second quarter of 2026, a year-on-year increase of 4.4%. Our non-e-commerce marketing services continued to expand in the content consumption, local life services, and AI application sectors, while our e-commerce marketing services demonstrated resilience under initiatives for integrated omnichannel traffic and brand growth.
Simultaneously, by applying AI technology, we have continuously deepened its integration across the entire marketing service workflow. In the second quarter of 2026, the content consumption, local life services, and AI application sectors drove sustained year-on-year growth in non-e-commerce marketing revenue.
In the content consumption sector, AI has reduced content production costs and lowered creative barriers, driving a rapid expansion in the supply of short dramas. This has catered to more diverse user preferences, further enriching the platform's content ecosystem and stimulating related marketing demand. As of June, the supply of short dramas on the Kuaishou platform, including both live-action and AI-generated short dramas, increased by more than fivefold compared to January.
Online marketing spend for short dramas in the second quarter grew by over 100% year-on-year. In the local life services sector, we continue to deepen our presence in niche markets such as local services in China, exploring incremental growth opportunities. We have also enhanced our full-funnel lead promotion products by optimizing deep conversion capabilities and launched a audience exploration agent to help clients more effectively identify high-intent customers, thereby improving lead quality and subsequent conversion efficiency.
In the AI application sector, we collaborate with clients to improve consistency between ad delivery and in-app user reception, helping them enhance user retention and conversion outcomes. This further consolidates our competitiveness in capturing client budgets within the AI application sector.
In terms of e-commerce marketing services, we continue to deepen the integration of traffic across our e-commerce and commercialization businesses, striving to optimize the matching efficiency between traffic and merchants. On one hand, we implement refined tiered operations for merchants, aligning differentiated product strategies with the core demands of different merchant types.
On the other hand, we have addressed the supply side of creative assets by actively implementing governance measures. By supporting first-release content to enhance recommendation diversity, we have effectively adjusted the structure of e-commerce marketing assets, enabling high-quality content to reach target traffic more effectively.
Meanwhile, we continue to optimize the long-term commercial ecosystem. Despite ongoing challenges in macro consumption and merchant operations, we remain committed to providing traffic support to high-quality merchants. The T2000 Brand Special Project, launched in the fourth quarter of last year, has shown initial success. Marketing spend by brand merchants grew faster than the overall e-commerce marketing service segment in the second quarter, contributing to a continuous increase in the share of online marketing revenue.
At the product level for e-commerce marketing, our Net GMV ROI products have undergone continuous upgrades. By enhancing net GMV bidding capabilities across the full link and all scenarios, and iterating bidding and auction strategies, customer penetration increased from 45% in the first quarter to 55% in the second quarter, effectively improving merchants' advertising ROI.
In the second quarter of 2026, we continued to optimize the application of AI in specific industry scenarios, further improving the efficiency of customer marketing campaigns and enhancing our capacity to absorb industry marketing budgets. In content consumption scenarios, AI facilitates intelligent targeting through content understanding and audience matching, helping high-quality content reach users more efficiently.
In local life service scenarios, AI capabilities are primarily applied to marketing asset generation, conversational live streaming operations, audience intent recognition, and deep conversion forecasting, helping merchants reduce costs related to content creation, campaign operations, and customer acquisition.
Fourth, regarding e-commerce business, in the second quarter of 2026, we maintained our strategic focus on buyer growth, supply introduction, and the deep integration of e-commerce and commercialization traffic. We optimized the merchant ecosystem structure, strengthened the introduction and cultivation of brands and new merchants, and promoted synergistic efficiency gains between e-commerce and commercialization.
In the second quarter of 2026, we focused on the growth of high-quality buyers, with the number of active e-commerce buyers remaining largely stable quarter-over-quarter. As users gradually form cross-domain consumption habits on the platform, we are strengthening our private domain advantages while connecting traffic, content seeding, purchase conversion, and store repurchases across various domains, gradually forming a positive cycle.
At the same time, we further improved synergy efficiency across domains and subsidy efficiency, achieving balanced development between content-driven and shelf-based commerce. On the e-commerce supply side in the second quarter of 2026, we continued to advance the introduction of new merchants and brand expansion. By empowering operations through cost reduction, efficiency enhancement, and growth incentive products, we continuously supported the growth of new and small-to-medium-sized merchants, further improving the merchant ecosystem structure.
We launched an upgraded version of the 'Star Shine Program,' providing tiered and targeted support to different groups, including major brand merchants, industrial belt merchants, and small-to-medium-sized new merchants, to help more merchants accelerate their scale growth. Driven by these policies, the number of newly onboarded merchants achieved year-over-year positive growth and nearly 10% quarter-over-quarter growth in the second quarter. The GMV from new merchants increased by nearly 30% year-over-year, with steady improvements in the quality of new merchant growth.
Branded merchants maintained a high year-over-year growth rate in the T2000 brand-specific initiative, with their share of the overall platform GMV continuing to rise. Brand-related commercial spending has achieved rapid year-over-year growth, and brands' dual contribution to platform GMV and commercialization continues to increase.
From an industry perspective, leveraging the advantages of content-driven e-commerce, the number of merchants in sectors such as tea, alcohol, health, beauty, and fresh food has continued to grow, creating structural growth opportunities. In the second quarter of 2026, we continued to optimize the creator ecosystem structure and improve the quality of content supply. By deepening cooperation with top-tier creators and increasing support for mid-tier creators in Kuaishou's advantageous verticals such as agriculture, rural areas, farmers (San Nong), and ACG (Anime, Comics, and Games), we stabilized the performance of existing creators, providing strong support for the foundational base of e-commerce content.
By integrating the platform's high-quality creator resources with distinctive product assortments from various regions across the country, we delved into local industrial clusters nationwide. We launched specialized content marketing initiatives, such as source-tracing live streams, to strengthen the synergy between content and supply, helping creators improve their efficiency in merchant integration.
Meanwhile, we continued to expand creator acquisition through internal incubation partnerships and external introductions. Regarding creator activity levels, we continuously iterated our attendance policies. In the second quarter, the number of effective live streams by 'Ai Fen Jia' anchors maintained net year-over-year growth, and creator broadcasting consistency steadily improved.
In building our distribution system, we leveraged AI technology to underpin core product capabilities, establishing a precise distribution framework that has led to a thriving distribution ecosystem. In the second quarter, the penetration rate of active distributors among creators continued to rise year-over-year, and the number of matches between products in the distribution library and creators increased by over 20% year-over-year.
During the second quarter, focusing on the full lifecycle of merchant operations, we continuously optimized AI capabilities across various e-commerce scenarios to help merchants reduce costs and increase efficiency, driving intelligent business operations. This validated the potential for AI to evolve from a tool for efficiency improvement to one that directly supports operational actions. In the first half of 2026, over 850,000 merchants on the platform used free AI business tools provided by the platform, covering multiple scenarios including product selection, listing, material production, business diagnostics, smart advertising, and customer service, offering end-to-end operational assurance and capability support.
Looking ahead, in the live streaming business, revenue reached RMB 8.7 billion in the second quarter of 2026. We focused on the healthy development of the supply side, combining AI-enabled live streaming products to continuously promote ecosystem quality improvement and product innovation.
On the supply side, we launched the 'Baichuan Plan,' which drove steady growth in new live streaming supply from guilds through user acquisition incentives, cold-start support, and healthy governance measures, thereby improving the cold-start efficiency for new streamers. Meanwhile, we strengthened the operation of independent streamers, focusing on identifying high-value independent talent and using refined operations to solidify the foundational base of live streaming supply.
Furthermore, we encouraged top-tier streamers to consolidate the group broadcasting format, leveraging their traffic and influence advantages to enrich the supply of high-quality live streaming content. On the product and technology front, AI capabilities further empowered live rooms. Based on potential AI video generation capabilities, we introduced customizable live-stream-exclusive effects. AI gifts continued to iterate, with more diverse gift formats and generation capabilities, enhancing users' willingness to pay.
In the second quarter of 2026, users sent over six million AI-driven gifts. AI-driven content understanding capabilities continuously optimized live streaming recommendation strategies, strengthening precise matching between streamers and users, thereby helping to increase the scale of paying users.
Furthermore, the live-streaming intelligent gift recommendation and ranking feature, based on real-time multimodal signals, has effectively enhanced user payment experience and efficiency. Functions such as the AI interactive assistant and digital avatar services continue to undergo iterative optimization, improving hosts' service efficiency.
Finally, regarding the progress of our overseas business, in the second quarter of 2026, we firmly executed our high-value growth strategy, solidifying the foundation of our overseas operations in terms of long-term profitability and localized深耕 (deep cultivation). In terms of traffic growth and content ecosystem, we adhered to refined user acquisition, strengthened the network of localized featured content and community-based creators, created a vibrant community atmosphere centered around real-life scenarios, and drove the deepening of core users' content consumption.
In the overseas online marketing business, leveraging major events such as the June Festival and the World Cup, we utilized AI-driven insights, relying on in-depth audience research, innovative product mechanics, and coordinated industry strategies to help marketing clients achieve rapid growth during key windows. Additionally, the monetization potential of diverse content forms, such as short dramas, continues to be unleashed, forming a dual-engine drive with our marketing service product capabilities, while accelerating expansion into incremental industries such as e-commerce.
Regarding the overseas e-commerce business, GMV and order volume continued to grow steadily year-over-year in the second quarter. Through optimization of merchandise structure and high-quality supply, we drove an increase in average revenue per user (ARPU), maintaining robust operational efficiency and profit quality.
In summary, looking ahead to the second half of the year, despite increasingly severe external challenges, we will remain steadfast in our core AI strategy, leveraging our technological and ecological assets to break through barriers and move forward. Although we are currently experiencing short-term pain points due to revenue pressure and significant AI investments impacting profit levels, we will adhere to long-termism to continuously expand the commercial boundaries enabled by AI. We aim to empower Kuaishou's content and commercial ecosystems, striving for both short-term breakthroughs and long-term high-quality growth, thereby creating long-term value for our vast user base and partners.
That concludes my sharing. Thank you all. Next, please welcome the Chief Financial Officer to introduce the company's financial performance for the second quarter of 2026.
Jin Bing
Thank you, Yixiao. Hello everyone. In the second quarter of 2026, we continued to deeply integrate AI technology into various business scenarios, achieving high-quality growth across our overall business amidst a complex environment. As the engine driving the company's long-term development, the comprehensive application of AI has not only deeply empowered the content ecosystem and enhanced user experience but also provided merchants and advertisers with full-link intelligent operation tools. This has reduced costs and increased efficiency for platform partners while injecting new growth momentum into our business.
In the second quarter of 2026, the Group's total revenue reached RMB 35.5 billion, with adjusted net profit amounting to RMB 3.9 billion. The adjusted net profit margin stood at 11%, indicating that the Group's overall profitability remains at a healthy level.
Next, let us take a detailed look at the company's financial performance for the second quarter of 2026. Total revenue for the second quarter of 2026 reached RMB 35.5 billion, representing a year-over-year increase of 4.1%. This revenue growth was primarily driven by the growth of our online marketing services and potential AI-related businesses.
Online marketing services revenue in the second quarter reached RMB 20.6 billion, a 4.4% increase from RMB 19.8 billion in the same period of 2025. This growth was primarily driven by the deepened application of AI in online marketing service processes, which effectively improved clients' advertising efficiency and subsequently boosted ad spending by marketing clients.
Other services revenue in the second quarter, including e-commerce and potential AI-related income, amounted to RMB 6.2 billion, an 18.5% increase from RMB 5.2 billion in the same period of 2025. This was mainly attributable to the growth of potential AI businesses. With breakthroughs in AI model capabilities, improvements in product features, and deeper integration into professional application scenarios, commercial monetization has maintained rapid growth.
Live streaming business revenue in the second quarter was RMB 8.7 billion. We continue to strengthen our advantage in content supply, leverage AI to empower product innovation, and promote ecosystem quality improvement, aiming to build a rich and healthy live streaming ecosystem with diversified content.
In terms of costs, cost of revenue in the second quarter increased by 10.7% year-over-year to RMB 17.2 billion, accounting for 48.4% of total revenue. This increase was mainly due to higher revenue-sharing costs and related taxes associated with revenue growth. Based on the above, the company's gross profit in the second quarter was RMB 18.3 billion, compared to RMB 19.5 billion in the same period last year. The gross margin was 51.6%, down from 55.7% in the same period last year.
Regarding expenses, sales and marketing expenses in the second quarter were RMB 9.9 billion, compared to RMB 10.5 billion in the same period last year. As a percentage of total revenue, this decreased from 30% in the same period last year to 27.9% in the current quarter, primarily due to reduced spending on promotional activities.
R&D expenses increased by 34.7% year-over-year to RMB 4.6 billion, accounting for 12.9% of total revenue. The rise in R&D expenses was mainly due to increased investment in AI, including related training costs. Administrative expenses were RMB 895 million, compared to RMB 897 million in the same period last year, remaining largely stable.
In the second quarter of 2026, the Group's net profit was RMB 3.2 billion, and adjusted net profit was RMB 3.9 billion, with an adjusted net margin of 11%.
Our balance sheet remains very strong. As of June 30, 2026, total available funds, including cash and cash equivalents, time deposits, financial assets, and restricted cash, amounted to RMB 121.3 billion. Net cash generated from operating activities in the second quarter of 2026 reached RMB 5.9 billion.
We actively execute our shareholder return strategy in a timely manner based on market conditions. As of August 19, the company had cumulatively repurchased approximately HKD 1.97 billion worth of shares during the year, buying back about 43.3 million shares, representing approximately 1% of the total share capital at the beginning of the year.
Looking ahead to the second half of the year, as Yixiao mentioned, facing external challenges, we expect to experience short-term pain characterized by revenue pressure and an impact on profit levels due to our steadfast and continuous investment in AI. We will continue to adhere to the philosophy of being technology-driven and user-first. While deeply addressing user needs, we remain committed to AI investment, leveraging leading AI technology to deeply empower the content ecosystem and commercial value chain. By maintaining prudent financial discipline to reduce costs and improve efficiency, we aim to further strengthen our core competitive barriers and create long-term value for users, partners, and shareholders.
That concludes our presentation. We will now move to the Q&A session. Thank you.
Operator
A quick reminder: if you wish to ask a question via phone, please press *1. The conference moderator will accept your request to speak; please hold. The first online question comes from Kenneth Fong at UBS.
Kenneth Fong
Thank you to management for taking my question, and congratulations on having such a robust core business alongside the strong growth potential of Keke AI. My question concerns the competitive landscape and iterative direction of Keke AI. Recently, we have observed timely updates from many video generation models. How should we view the competitive dynamics in this space, and what will be Keke AI's competitive strategy?
Cheng Yixiao
Thank you for the question. The global market for video content generation is a massive $150 billion sector, where AI video generation models hold significant penetration potential. We believe the AI video generation track is currently in a prosperous phase characterized by diverse participation driving industry growth. Numerous players are leveraging their platform ecosystems, vertical scenarios, and technical strengths to engage in differentiated competition, continuously expanding boundaries across advertising, e-commerce, film, short dramas, gaming, and other commercial applications.
However, unlike the relatively fragmented competitive landscape of large language models, the AI video generation sector currently features clear leadership and high concentration among the top tier. Video generation models impose higher requirements on computing power, data, talent, and technology.
Over the past two years, Keke AI has firmly maintained its position in the first tier of the AI video generation sector, with the team consistently demonstrating forward-looking strategic judgment and execution capabilities. In June 2024, we launched the world's first commercially viable video model product based on the DiT architecture. In April 2025, we released the world's first Multimodal Visual Language (MVL) interaction mechanism. In December 2025, we introduced the world's first Omni-model for multimodal video generation.
We continue to demonstrate through our achievements that the Keke AI team is a top-tier global team with strong capabilities in both research and engineering, as well as strategic execution. Recently, Keke AI completed independent financing, which will further enhance its competitiveness in the industry.
Since its inception, Keke AI has focused on serving professional content creators, using large video generation models to improve the efficiency of professionals who rely on video creation as their career and have sustained payment capacity. Keke AI possesses strong prompt understanding and controllable storyboard capabilities, helping users achieve precise and coherent creative expression.
In terms of image quality and production scalability, Keke AI also meets the demand of professional creators for high-quality video content. Keke AI is the world's first video model capable of native 4K output, allowing users to generate 4K videos with a single click without the need for additional upscaling or post-production enlargement. This results in clearer visuals, richer details, and a more cinematic texture.
In summary, we are confident in the long-term competitiveness of Keling AI in the video generation sector. As model capabilities continue to iterate, product experience improves, and the ecosystem for professional creators and commercial applications expands further, Keling AI will continue to unlock its growth potential during the rapid penetration of the AI video generation industry.
Operator
The next question comes from Lincoln Kong at Goldman Sachs.
Lincoln Kong
Thank you, management, for taking my question. I would like to ask about the AI strategy. Apart from Keling AI, what progress has been made in other AI areas this quarter?
Cheng Yixiao
No problem. This quarter, the company's progress in the AI field is not only reflected in the continuous iteration of Keling AI but also includes substantial achievements in enhancing organizational efficiency, implementing online marketing service scenarios, and iterating the core recommendation system.
Regarding the implementation of online marketing services, we have deeply integrated AI into key processes such as AIGC material generation, intelligent bidding, and generative recommendations. On the material generation side, by connecting user interest modeling with video production, we have upgraded from merely sourcing marketing video materials to creating videos tailored for users.
On the intelligent bidding side, the agent system combines the historical data of marketing clients' ad accounts with their conversion goals to provide self-learning, continuously corrected automated delivery strategies. This effectively safeguards clients' long-term value and campaign performance, yielding strong returns.
Generative recommendations enable the model to truly understand ad content and user needs, translating information from live commerce, search, and industry sectors into semantic expressions, thereby improving the efficiency of matching people with products.
In terms of recommendation system iterations, we launched Agent X, an agent-driven closed-loop R&D product. Previously, the process of rolling out recommendation strategies—from idea conception to launch—involved multiple steps, including data analysis, solution design, coding, experimental configuration adjustments, A/B testing, metric observation, attribution, and review. This process heavily relied on manual efforts by algorithm engineers, limiting efficiency.
Today, agents have become the primary executors of recommendation iterations, freeing engineers from repetitive tasks and allowing them to focus on goal setting, key reviews, and high-level judgment. This has significantly improved both the iteration efficiency and the final outcomes of recommendation strategies.
In terms of organizational efficiency improvement and empowerment, the company's self-developed agent products, including My Flicker, have achieved an employee penetration rate of over 90%. These tools comprehensively cover multiple departments such as technology R&D, data analysis, and business operations, significantly boosting internal efficiency.
Taking technology R&D efficiency as an example, in Q2 2026, the average requirement delivery cycle for R&D technical staff shortened by more than 10% compared to Q1. Meanwhile, the average daily lines of code submitted by R&D personnel with AI assistance increased by over 70%, and the AI code contribution rate has surpassed 60%.
Kuaishou Wanqing not only efficiently supports Kuaishou's internal AI usage scenarios but also provides large-model infrastructure services to numerous external corporate clients, achieving solid revenue growth.
In summary, we will continue to deeply empower our business and organization with AI technology, constantly expanding the boundaries of application. Working hand in hand with our broad partner base, we will jointly explore and create commercial value and growth opportunities with greater imagination.
Operator
The next question comes from Thomas Chong at Jefferies.
Thomas Chong
Good evening, and thank you to the management team for taking my questions. Against a backdrop of uncertain macroeconomic demand and industry competition, how do we alleviate pressure on merchants and support their growth? Additionally, what is our outlook on e-commerce development trends for the second half of the year?
Cheng Yixiao
Thank you for your question. Macroeconomic demand remained under pressure in Q2, and industry competition continued to be intense. In this environment, merchants prioritize operational certainty, which serves as the starting point for our merchant-focused initiatives. We implement tiered management based on merchants' business types, aligning organizational policies and resources to ensure that different types of merchants can find growth paths suited to their needs on Kuaishou.
First, focusing on merchants' business types, we adjusted the division of labor within our supply team starting in Q2. We matched differentiated product strategies to the core demands of different merchant types and restored policy resources to more reasonable levels.
Previously, there was a mismatch between the policies and resources allocated to different merchant types and their contributions and demands. Brand merchants lacked sufficiently professional services; advertising-driven merchants received e-commerce resources that were disproportionately low relative to their revenue contributions; while persona-based and content-driven merchants, who enjoyed abundant traffic resources, primarily needed support for sustainable operations.
To this end, we have consolidated brand merchants under specialized teams for unified management. We are reallocating mismatched e-commerce resources away from ad-driven merchants, while focusing on reaching fans across all channels for persona-based and content-driven merchants to help enhance the sustainability of their operations.
Building on this foundation, we continue to provide tiered and targeted support to major brand merchants, industrial belt merchants, and small and medium-sized enterprises (SMEs). We aim to maintain policy stability and continuity to ensure more precise allocation of resources.
AI is a key lever for helping merchants reduce costs and improve efficiency. AI capabilities have been deeply integrated into daily operations and cost management across scenarios such as intelligent content production, smart marketing, automated after-sales service, and intelligent customer support, enabling merchants to reinvest savings into genuine growth.
Brands and industrial belts are our two strategic focus areas on the supply side this year. Since the fourth quarter of last year, tighter compliance policies have leveled the compliance costs across different merchant types. As merchants demand higher ROI on traffic acquisition, the advantages of brand merchants in operational capability and stability are becoming more evident.
Furthermore, with AI significantly lowering the barrier to content creation and our brand subsidies being more focused on core hit products, brand merchants on Kuaishou have entered a virtuous cycle. We observe these brands establishing a solid presence on the platform, focusing on substantive operations rather than mere exposure or external traffic redirection.
In the second quarter, both the proportion of brand self-operated sales in the overall GMV and the share of advertising spend by brand merchants achieved stable growth. On the industrial belt front, we leverage service providers to deepen engagement with industrial clusters nationwide, empowering local merchants. This year, we have implemented these initiatives in regions such as Inner Mongolia, Yunnan, Shandong, and Jiangsu, effectively delivering operational policies and growth incentives to help regionally distinctive merchants establish roots on Kuaishou and achieve scaled operations.
Regarding e-commerce trends for the second half of the year, we assess that the broader consumption market is still in a phase of moderate recovery, with the shift from discretionary to essential consumption continuing. This implies that merchants will increasingly demand operational certainty, and the growth model relying solely on traffic dividends is no longer viable.
While we continue to provide merchants with traffic resources and commission rate support, we anticipate that e-commerce marketing service revenue and commission income will face year-on-year pressure in the second half of the year. For the platform, this presents both challenges and opportunities.
We will continue to steer our traffic integration mechanisms toward brands and balanced merchants with the capability and willingness for long-term operations. We will also deepen our intelligent advertising capabilities to help merchants improve the certainty of their ad spending. We believe that by solidifying tiered merchant management and optimizing the supply structure, our e-commerce business can achieve high-quality growth amidst structural opportunities. This is essential for us to navigate cycles and achieve long-term sustainable development.
Operator
The next question comes from Daniel Chen at JP Morgan.
Daniel Chen
Thank you to the management team for the opportunity to ask questions. I would like to inquire about the development trends in non-e-commerce online marketing services, specifically advertising, for the second half of the year. Additionally, what has been the enabling effect and actual impact of AI capabilities on the online advertising business this quarter?
Cheng Yixiao
Thank you for your question. The growth of online marketing services is influenced by the macroeconomic environment and clients' marketing budgets. Looking ahead to the second half of the year, we believe visibility into the external environment remains limited. Meanwhile, while AI applications and the instant retail sector saw rapid growth in the second half of last year, they face a high base effect in the second half of this year. Client budgets are also affected by the pace of industry competition and advertising placement strategies.
However, we also observe certain structural opportunities within the industry. First is content consumption represented by short dramas. We believe this industry has not yet reached its growth ceiling. AI continues to reshape the supply side of the short drama industry, significantly reducing production costs and turnaround times, thereby bringing more themes, diverse content formats, and richer supply.
In 2025, Kuaishou hosted 60,000 short drama series. We expect this number to reach 500,000 by 2026. Going forward, industry development will gradually shift from quantitative expansion to improving content quality. At the same time, we see the value of the short drama content ecosystem becoming increasingly prominent.
By July 2026, the daily average exposure volume for Kuaishou short drama users had reached 230 million. The monetization model for short dramas is also evolving, with an increasing share coming from In-App Advertising (IAA). This indicates that short dramas are gradually developing into an ecosystem business that possesses both content consumption value and advertising承载 capacity.
Furthermore, future content formats may extend beyond simple viewing towards IP-derived interactive dramas, virtual companionship, and the integration of culture, tourism, and creative industries. These developments could bring new opportunities for content consumption and commercialization. In line with these trends, we will seize opportunities in content consumption development by helping high-quality content achieve more stable traffic and commercial returns through revenue-sharing incentives, platform-wide traffic support, and premium content copyright collaborations.
Second, our penetration in certain niche sectors has been insufficient, leaving room for further growth. For example, in the local life services sector, which is highly correlated with the macroeconomic environment, there are numerous sub-sectors. Merchants' operating models, user decision-making journeys, and conversion goals vary significantly across different industries.
Therefore, we will continue to uncover growth opportunities in vertical tracks such as beauty and wellness, home improvement, and real estate through more refined industry operations. In the gaming industry, while game content has a substantial user consumption base on the Kuaishou platform, commercialization remains relatively weak.
Looking ahead, we will explore opportunities to integrate game live-streaming content with advertising budgets. Furthermore, in the AI era, the barrier to entry for developing mini-games has been further lowered, which may also drive new supply of mini-games and marketing demand.
Regarding AI's empowerment of online marketing services, as clients increasingly focus on operational efficiency and the ROI of marketing spend, AI has become a key tool for us to help clients improve efficiency, reduce costs, and unlock their marketing budgets.
Specifically, first, in terms of marketing creatives, AI has significantly lowered production barriers. For instance, in sectors such as local life services using AI tools, AI helps clients and service providers produce content and materials suitable for Kuaishou users more quickly and at a lower cost, thereby driving growth in marketing demand across related industries.
Second, in client operations, AI is being deeply integrated into more workflows, including business opportunity insights, product selection recommendations, ad placement diagnostics, performance reviews, and customer support. This helps clients and service providers standardize and automate processes that were previously heavily reliant on manual labor, thereby enhancing operational efficiency.
Meanwhile, on the platform side, we are continuously leveraging AI models to improve the matching and conversion efficiency of our advertising system. For example, through better creative understanding, user interest profiling, smart bidding, and dynamic pricing capabilities, we enable online marketing clients to allocate their budgets more efficiently to the right audiences and scenarios.
In summary, there remain structural growth opportunities in online marketing services in the second half of the year. AI will continue to serve as a critical capability for enhancing marketing efficiency and delivery returns, driving improvements in client operational efficiency and commercial growth.
Operator
The final question comes from Yuan Liao of Citi.
Yuan Liao
Good evening, management. Thank you for taking my question. I would also like to congratulate you, as Keai (AI) has recently completed independent financing. My question concerns the company's cash flow. How will Keai's completion of financing impact Kuaishou's cash flow? Additionally, does the company have any plans for cash outflows in the second half of this year? What is the outlook for our cash flow situation in the second half?
Jin Bing
After completing its financing, Keai (AI) will adopt more flexible approaches to cash flow support, including addressing greater computing power needs through leasing and other methods, thereby further optimizing capital allocation. This will, to some extent, improve the Group's overall cash flow position.
We will continue to adhere to a prudent financial strategy in terms of cash management and capital expenditures. In terms of execution pace, the majority of capital expenditures were concentrated and completed in the first half of the year. As a result, the company's free cash flow was positive in the second quarter, and our goal is to achieve positive group-level free cash flow in the second half of the year.
Regarding shareholder returns, we remain committed to creating long-term value as our core objective and continue to implement an active return policy. Since the beginning of this year, the company has completed approximately HK$2 billion in share buybacks and HK$3 billion in cash dividends, bringing total shareholder returns close to the full-year level of last year.
Against the backdrop of firmly advancing our AI strategic investments, we have still achieved a steady increase in the intensity of shareholder returns. We expect total shareholder returns for the full year 2026 to surpass last year's levels. These measures not only provide tangible returns to shareholders but also reflect the company's ability to consistently generate cash flow and its confidence in the long-term development of the business.
While fully embracing the AI era, the company will continue to fortify its financial safety baseline, reduce costs and improve efficiency through prudent financial discipline, and maintain ample healthy cash reserves. We will use a robust financial structure as our ballast to achieve a healthy balance between business expansion and shareholder returns, laying a solid foundation for high-quality long-term growth.
Operator
That concludes today's Q&A session. Now, please welcome Matthew for his closing remarks.
Matthew Zhao
Thank you all for participating today. If you have any further questions, please feel free to contact Kuaishou's Capital Markets and Investor Relations team at any time. Thank you.
More details:KUAISHOU-W IR
Disclaimer: The above content is generated by an AI language model based on public data and third-party automatic subtitles. The above content does not represent any position of Futu and does not constitute any investment advice. Futu Group makes no express or implied warranties or representations regarding the accuracy, timeliness, or completeness of the above content.
Risk Disclaimer: The above content only represents the author's view. It does not represent any position or investment advice of Futu. Futu makes no representation or warranty.Read more
Comment (1)
to post a comment
152
20