Q2 2024 Baidu Inc Earnings Call
Speaker Change: hello and thank you for standing by for bo second quarter and fiscal year two ythousand and twenty four earnings conference call
Operator: quarter and fiscal year 2024 earnings conference call.
Operator: At this time all participants are in listen-only mode.
Speaker Change: at this time all participants are in listen only mode after management's prepared remarks they will be a question and answersession today's conference is being recorded if you have any objections you may disconnect at this time
Operator: After management's prepared remarks, there will be a question and answer session.
jimline by: i would now like to turn the meeting over to your host for today's conference jimline by who is director of investor relations
Speaker Change: hello i and welcome to by the second quarter two thousand and twenty four earnings compference call by these earnings release was distributed earlier today and you can find a copy on our wbside as well as our new to our services
Speaker Change: and the call today we have roubbin me our cor per and ceo growing more our cfo and adoion our p in charter by to ai coud group dg after our prepared remarks we will hold a to an session please note the discussion today we will contain forward-looking statements made under the safe public provisions of the u s credit security litigation reform act of onethousandnine nety five
Speaker Change: forward-looking statements are subject to risk and uncertainties that may cause actual results to differ mataterurially from our contte like expectations
Speaker Change: for detailed discussions of these risks and uncertainties please refer to our relatest to annual report and other documents cloued with sec and h ong exchange
Speaker Change: but i do does not undertake any obligation to asthink any forward-looking statements except as required under applicable law
Speaker Change: our earnings press release and this call include discussions of certainen and not non-gaap financial measures
Speaker Change: our press release contains a reciliation of the annic and-gaap measures
Operator: Today's conference is being recorded.
Speaker Change: to the announed in most therdireaply compfromable gaap measures and is available on our ir website at ir by do com as a reminder this conference is being recorded in theaddition over acost of this conference call will be available on by this irour website
Operator: If you have any objections, you may disconnect at this time.
Robin: i will now turn the call over to our ceo robin
Operator: I would now like to turn the meeting over to your host for today's conference, Gem Lin, Baidu's Director of Investor Relations.
Robin: hello and one byital coreter for our revenue growth slightly to rb twenty-six point seven billion in q two thanks to the continued acceleration of our ni cld business despite the headwinds in our online margin bins
Gem Lin: Hello everyone and welcome to Baidu's second, quarter 2024 earnings conference call.
Gem Lin: Baidu's earnings release was distributed earlier today and you can find a copy on our website as well as our newsletter services.
Gem Lin: On the call today we have Robin Li, our co-founder and CEO, Rong Luo, our CFO, and Dou Shen, our EVP in charge of Baidu AI Cloud Group, ATG. After our, prepared remarks, we will hold a Q&A session.
Speaker Change: even with the ongoing investment in ai by the co non-gaap operating margin improved by close to two percentage pots year-over-year to twenty six percent and non-gaap operating profit grow by eight percent year-over-year
Gem Lin: Please note that the discussion today will contain forward-looking statements made under the safe harbor provisions of the U.S. Credit Security Litigation Reform Act of 1995.
Gem Lin: Forward-looking statements are subject to risks and uncertainties that may cause, actual results to differ materially from our current expectations.
Gem Lin: For detailed discussions of these risks and uncertainties, please refer to our latest annual report and other documents filed with the SEC and Hong Kong Stock Exchange.
Speaker Change: sincese to the operational efficiency gains we continue to implement
Speaker Change: by the near-term pressure we remain fully competidenting our strategic direction and the transformation transform potential of geni and foundition models
Gem Lin: Baidu does not undertake any obligation to update any forward-looking statements except as required under applicable law.
Speaker Change: as we move further in into two thousand and twenty four we are happy to observe a significant change as i do is skating a i to address real world problems and generate substantial value both externally and internally
Gem Lin: Our earnings press release and this call include discussions of certain unnoticed non-GAAP financial measures. Our press release contains a reconciliation of the unnoticed non-GAAP measures to the unnoticed and most directly comparable GAAP measures and is available on our IR website at ir.baidu.com.
Speaker Change: central to this effort is our commitment to making earnings increasingly affordable and accessible
Operator: As a reminder, this conference is being, recorded.
Operator: In addition, a webcast of this conference call will be available on Baidu's IR website.
Speaker Change: i like to highlight some keing top our progress
Speaker Change: with the scaling of ai is accelerating at a britneck pace relacking the real value that earn creates for people and for bin ali
Gem Lin: I will now turn the call over to our CEO, Robin.
Robin Li: Hello everyone.
Speaker Change: we find that the most tangible benefit from ai comes from the adoption and you ofckt applications bue on top las
Robin Li: Baidu Core's overall revenue grows slightly to RMB 26.7 billion in Q2. Thanks to the continuous acceleration of our AI cloud business despite the, headwinds in our online marketing business. Even with the ongoing investment in AI, Baidu Core's non-GAAP operating margin improved by close to two percentage, points year-over-year to 26% and non-GAAP operating profit grew by 8% year-over-year. Thanks to the operational efficiency gains, we continue to implement.
Robin Li: Despite the near-term pressure, we remain fully confident in our strategic direction and the transformative potential of Gen AI and foundation models.
Speaker Change: just three months ago we announced that earie and no about two an a million api cost daily recently its thirdpast six hundred million or over one trillion tokens are generated every day
Speaker Change: to make earning more affordable we continue to expand our model potapfornial and enhance our model capabilities to meet diverth customer needs
Speaker Change: last quarter we launched a sweet like we earning models and they have quickly gained attraction among enterprises and developers
Speaker Change: building all this momentum we introduced the earning four point zero turable in june it offers superior capabilities compared to earning four point zero for typical usecases yet is designed to be much cheaper and faster to wr
Speaker Change: are conlylineing up now includes our flagship models earnings three point five and earnings with four point zero and the enhenced the model earning four point zero turable and several lightweight models
Speaker Change: this diverse portfolio allows us to accommodate the varying needs of our customers optimizing for performance cost and latency
Speaker Change: in maye we made a strategic move to make avi callst free of charge for three lightweight earning models that's earning speed early light and earning timing
Speaker Change: and in july we significantly lowered the prent of api half for the two earning flle sha models earnings three point five and earning four point zero
Speaker Change: this decision is rooted in our commitment to enabling wider access to earning and making ai accessible to all
Speaker Change: at the same time we continue to lower the cost of model inference one optimization of note this quarter it's the upgrade of pedo pedtal our opensource if learn new framework to a version sthree point zero
Speaker Change: this upgrade significantly improves the framework's compatibility with our aitinfrastructure and learning which we expect well further help toreduce model inference costs in the future
Robin Li: As we move further into 2024, we are happy to observe a significant change as Baidu is scaling AI to address real-world problems and generate substantial value both externally and internally.
Robin Li: Central to this effort is our commitment to making earnings increasingly affordable and accessible.
Speaker Change: i briefly mentioned earlier that by du scales ai to address world problems and generate substantial value both externally and internally i'd like to discuss this in more detail
Robin Li: I'd like to highlight some key points of our progress.
Robin Li: The scaling of AI is accelerating at a breakneck pace, reflecting the real value that Ernie, creates for people and for business alike.
Speaker Change: extraally we have empowered our ai cloud customers to achieve greater efficiency and scalability by using learning
Speaker Change: our solutions have enabled clients to optimize their operations and realize significant benefits across whereious initiatives
Robin Li: We find that the most tangible benefit from AI comes from the adoption and use of applications, built on top of LLMs.
Speaker Change: let me offer use some useth cases to illustrate that real bv problems we have addressed and how we add addvalue to our customers
Robin Li: Just three months ago, we announced that Ernie handled about 200 million API calls daily. Recently, it surpassed 600 million, or over 1 trillion tokens are generated every day.
Robin Li: To make Ernie more affordable, we continue to expand our model portfolio and enhance, our model capabilities to meet diverse customer needs. Last quarter, we launched three lightweight Ernie models, and they have quickly gained, traction among enterprises and developers.
Robin Li: Building on this momentum, we introduced Ernie 4.0 Turbo in June. It offers superior capabilities compared to Ernie 4.0 for typical use cases, yet it's, designed to be much cheaper and faster to run. Our current lineup now includes our flagship models, Ernie 3.5 and Ernie 4.0, and the enhanced, model, Ernie 4.0 Turbo, and several lightweight models. This diverse portfolio allows us to accommodate the varying needs of our customers, optimizing, for performance, cost, and latency.
Robin Li: In May, we made a strategic move to make API calls free of charge for three lightweight, Ernie models, that's Ernie Speed, Ernie Light, and Ernie Tiny.
Robin Li: And in July, we significantly lowered the price of API calls for the two Ernie flagship, models, Ernie 3.5 and Ernie 4.0. This decision is rooted in our commitment to enabling wider access to Ernie and making, AI accessible to all. At the same time, we continue to lower the cost of model inference.
Robin Li: One optimization of note this quarter is the upgrade of PetalPetal, our open source deep, learning framework, to version 3.0. This upgrade significantly improves the framework's compatibility with our AI infrastructure and, Ernie, which we expect will further help reduce model inference costs in the future.
Speaker Change: in the health care industry doer burn out is's one of the greatest challenges
Robin Li: I briefly mentioned earlier that Baidu scales AI to address real-world problems and generate, substantial value, both externally and internally.
Speaker Change: by leveraging earning speed and our model buder we helped a health care automation solution provider to train and finetune and industryes spesaific model that enables automatically generated medical records for doctors
Robin Li: I'd like to discuss this in more detail.
Robin Li: Externally, we have empowered our AI cloud customers to achieve greater efficiency and, scalability by using Ernie. Our solutions have enabled clients to optimize their operations and realize significant benefits, across various industries.
Robin Li: Let me offer you some use cases to illustrate the real business problems we have addressed, and how we add value to our customers.
Robin Li: In the healthcare industry, Dr. Burnout is one of the greatest challenges.
Speaker Change: this solution significantly reducedces ces the administrative burden on doctorterss a major factor in burnout and enhensits clinical efficiencies
Speaker Change: after two months of models deployment doctors eieldingit were able to treat fifty percent more patient in on average
Speaker Change: in the recruitment industry we have collaborated with the recruiting service company by leveraging earnings capabilities school api h our customer has significantly upgraded the matching process between job ve cription and resues
Robin Li: By leveraging early speed and our model builder, we help a healthcare automation solution provider, to train and fine-tune an industry-specific model that enables automatically generated medical records for doctors. This solution significantly reduces the administrative burden on doctors, a major factor in burnout, and enhances clinical efficiency. After two months of model deployment, doctors using it were able to treat 50% more patients, on average.
Robin Li: In the recruitment industry, we have collaborated with a recruiting service company. By leveraging early capabilities through API calls, our customer has significantly upgraded, the matching process between job descriptions and resumes. The smart matching upgrade has reduced labor costs in this process by over 50%, measured, by total working hours, while maintaining high-quality results. This automation allows the customer to expand its services to a broader client base, thereby, increasing revenue-generating capability and demonstrating the tangible benefits of integrating earning into its operations.
Speaker Change: the smart matching upthgrade as we u the labor costs in this process by over fifty percent metered by total working hourlls while maintaining high-quality results
Speaker Change: this automation allows the customer to expand its servic es to a broader declient phase thereby increasing revenue generating capability and demonstrating the tangible benefits of integrating earning into its operations
Robin Li: The customer is very satisfied with the result and is now exploring new collaboration opportunities.
Speaker Change: the customerers is stair satisfied with a result and is's now exploring new collaboration opportunities
Robin Li: In the public service sector, we utilized early speed to deliver tailored, personalized, and scalable services.
Speaker Change: in the public service sector we utilized the earning speed to deliver tailored personalized and scalable servness advanting challenges like limited resources and efficiency needs
Robin Li: Improving challenges like limited resources and efficiency needs. This is particularly valuable in China's grassroots public services, where shortages in manpower, make large-scale, personalized services challenging. We have collaborated with the customer to help over 6,000 villages in China improve, public services, offering citizens more personalized and efficient support. Since the large-scale launch of this service in April, its daily usage has surged over, 15-fold to more than 2 million times per day, helping locals handle tasks such as household registrations, social security inquiries, and tax declarations.
Speaker Change: this is particularly valuable in china's grassroo public services where shortages in manpower makes large scale personalized the services challenging
Speaker Change: we have collaborated with the customer to help over six thousand villages in china improve topic services offering citizens more personalized and efficient support
Speaker Change: think the large scale launch of the servicing april its daily usage has searched over fifteen fold to morgthan two million times predday
Speaker Change: helping locals handle tasks suchas household regituations social security inquiries and tax declarations
Robin Li: Internally, we have accelerated the renovation of Baidu Search with Ernie, significantly, enhancing our consumer-facing products at large scale.
Speaker Change: internally we have accelerated the renovation not byid to search with earnie significantly enhancing our consumer feing products at large scale
Robin Li: Generative AI is transforming user search experience, pushing beyond traditional boundaries.
Speaker Change: generatedativeis ai is transforming user search experience pushing beyond traditional boundaries
Robin Li: Since the second half of last year, we have been testing early-powered features on a small, scale.
Speaker Change: since the second half of last year we have been testing earning powered features on a small scale this quarter our focusused has shifted to substantially amplifying search capabilities with advanced ai features
Robin Li: This quarter, our focus has shifted to substantially amplifying search capabilities with advanced, AI features. Currently, 18% of search result pages contain generated content, up from 11% in mid-May. This AI-generated search results deliver more accurate and direct answers, enhancing content, quality and providing previously unattainable information. This improvement has led to increased user satisfaction and engagement, as more users, are now turning to Baidu Search for more complex queries, increasing Baidu Search's versatility.
Speaker Change: currently eighty percent of search without pages contain generated the content
Speaker Change: up from eleven percent in midmay this ai generated the search without deliver more accurate and direct answers
Speaker Change: enhancing content quality and providing previously our attenable information
Speaker Change: this improvement has led to increased user sfaction and engagement
Speaker Change: at more users are now turning to by the search for more complex queries increasing by re searches versatility
Robin Li: AI-generated search results may reduce ad impressions and therefore have a negative, short-term impact on monetization, but they provide significant value to our users. By prioritizing long-term user experience over immediate revenue and profits, we see, the strategic implementation of generative search as essential to driving future success.
Speaker Change: again i generuated to search about may withredyou have impressions and therefore have a negative short-term impact on oneation but they provide significant value to our users
Operator: and Carter and fiscal year 2024 earnings conference call.
Operator: and Carter and fiscal year 2024 earnings conference call. At this time, all participants are in listen only mode. After management's prepared remarks, there will be a question and answer session. Today's conference is being recorded. If you have any objections, you may disconnect at this time.
Operator: At this time, all participants are in listen only mode.
Speaker Change: by prioritizing long-term user experienceced over immediate revenue and profits we see the strategic implementation of generative search as essential to viding future success
Operator: After management's prepared remarks, there will be a question and answer session.
Operator: Today's conference is being recorded.
Operator: If you have any objections, you may disconnect at this time.
Jim Lin: I would now like to turn the meeting over to your host for today's conference, Jim Lin, Baidu's Director of Investor Relations.
Robin Li: Our AI investments are fostering deeper user interaction. New interactive features enable users to refine their questions through multi-round conversations, enhancing the overall user experience.
Speaker Change: are aiu sment on our fostering deeper user inductions
Jim Lin: I would now like to turn the meeting over to your host for today's conference, Jim Lin, Baidu's Director of Investor Relations.
Jim Lin: Hello everyone and welcome to Baidu's second quarter 2024 earnings conference call. Baidu's earnings release was distributed earlier today and you can find a copy on our website as well as our new software services.
Speaker Change: new inter active features in users to refine their questions through m around conversations enhancing the overall user experience
Jim Lin: Hello everyone and welcome to Baidu's second quarter 2024 earnings conference call.
Robin Li: Additionally, there is a significant increase in users utilizing text and image creation, tools within search.
Jim Lin: Baidu's earnings release was distributed earlier today and you can find a copy on our website as well as our new software services.
Jim Lin: On the call today, we have Robin Lee, our co-founder and CEO Rong Luo, our CSO and adoption, our UVP in charge of Baidu AI Cloud Group ACG. After our prepared remarks, we will hold a Q&A session. Please note that the discussion today will contain forward-looking statements made under the safe proper provisions of the U.S. Credit Security litigation reform act of 1995. Forward-looking statements are subject to risk and uncertainties that may cause actual results to defer maturity from our context like expectations.
Speaker Change: additionally there is that significant increase in users utilizing text and image creation to us living search
Robin Li: This development opens doors for long-term value creation.
Speaker Change: this development open doors for long-term value credation
Robin Li: Last quarter, I mentioned that we plan to accelerate the distribution of earning agents. So far, we have seen more and more developers, advertisers, and partners gather on our platform, to develop innovative AI agents. Earning agent distribution within Baidu increased dramatically in July, exceeding 8 million, daily, more than double the number in May.
Jim Lin: For detailed discussions of these risks and uncertainties, please refer to our latest annual report and other documents filed with SEC and Hong Kong official exchange. Baidu does not undertake any obligation to update any forward-looking statements, except as required under applicable law. Our earnings press release and these call include discussions of certain and not-a-digment non-gap financial measures. Our press release contains a reconciliation of the anonymous non-gap measures to the and not-a-digment most directly comfortable gap measures. And it's available on our IR website at ir.bidu.com. As a reminder, this conference is being recorded. In addition, a webcast of this conference call will be available on Baidu's IR website.
Speaker Change: last quarter i mentioned that we plan to accelerate the distribution of earning agents
Speaker Change: so far we have seeing more and more developers advertiseersve and partners gather on our platform to develop innovative ai agents
Speaker Change: bernie agent distribution withining by the increased dramatically in july exceeding eight million daily more than double the number in may
Robin Li: Currently, the most frequently used agents by users are for content creation, personal, insights, such as personality testing, and tools such as translation and schedule planning.
Speaker Change: currently the most frequently used agents by users for counent quation personal insid such as personality testing and tools such as translation and schedule plan
Robin Li: At the same time, we also see a trend of earning agents being more widely developed, and applied, both internally and externally. Our first college application assistant agent is one example of how earning agents are being, distributed within Baidu to assist users with complex problem-solving and decision-making.
Speaker Change: at a same time we also was see a joint of earning agents being more widely divised and applied both internally and externally
Jim Lin: On the call today, we have Robin Lee, our co-founder and CEO Rong Luo, our CSO and adoption, our UVP in charge of Baidu AI Cloud Group ACG. After our prepared remarks, we will hold a Q&A session.
Robin Lee: I will now turn the call over to our CEO, Robin. Hello everyone. Baidu core for our revenue growth slightly to R&B 26.7 billion in Q2. Thanks to the continuous acceleration of our AI cloud service, despite the headlines in our online margin events. Even with the ongoing investment in AI, Baidu core's non-gap operating margin improved by close to two percentage points year-over-year to 26% and non-gap operating profit grow by 8% year-over-year. Thanks to the operational efficiency gains, we continue to implement.
Speaker Change: our first college application assistant agent is one example of howall earning agents are being distributed living by d to assist a users with complex proble solving and decision making
Jim Lin: Please note that the discussion today will contain forward-looking statements made under the safe proper provisions of the U.S.
Jim Lin: Credit Security litigation reform act of 1995.
Jim Lin: Forward-looking statements are subject to risk and uncertainties that may cause actual results to defer maturity from our context like expectations. For detailed discussions of these risks and uncertainties, please refer to our latest annual report and other documents filed with SEC and Hong Kong official exchange.
Jim Lin: Baidu does not undertake any obligation to update any forward-looking statements, except as required under applicable law.
Speaker Change: as a year after china's national college centinterest examination selecting uniit adversities after checking it them resuwithout is crucial for the ten million class candidates
Robin Li: Selecting universities after checking exam results is crucial for the 10 million plus, candidates. In June, Baidu launched an innovative college application assistant agent, which is designed, to meet personalized needs for selecting universities and majors. During the national college entrance exam, the peak daily active users of the agent, approached 2 million, highlighting its significant impact and utility.
Jim Lin: Our earnings press release and these call include discussions of certain and not-a-digment non-gap financial measures. Our press release contains a reconciliation of the anonymous non-gap measures to the and not-a-digment most directly comfortable gap measures. And it's available on our IR website at ir.bidu.com.
Speaker Change: in june bu launched an innovative college application assistant agent which is designed to meet personalized the needs for selecting universities and majors
Jim Lin: As a reminder, this conference is being recorded. In addition, a webcast of this conference call will be available on Baidu's IR website.
Speaker Change: following the national college entrance exam the peak dady active users of the agent approached two million highlighting its significant impact and utility
Robin Lee: I will now turn the call over to our CEO, Robin.
Robin Lee: Despite the near-term pressure, we remain fully competent in our strategic direction and the transformation, the transformating potential of Gen AI and foundation models. As we move further into 2024, we are happy to observe a significant change as Baidu is scaling AI to address real-world problems and generate substantial value both externally and internally. Thanks to this effort, it's our commitment to making early increasingly affordable and accessible. I'd like to highlight some key points on Baidu's The scaling of AI is accelerating at a breakneck pace, reflecting the real value that Ernie creates for people and for bins alike.
Robin Li: While all these new features for more sophisticated user needs have not yet been monetized, they, are a transformative force for search and are crucial for our future success.
Robin Lee: Hello everyone.
Speaker Change: well all these new features for more sophisticated user needs have not yet been mizeded they are transformative forth for search and crucial for our future with death
Robin Li: This strategic focus positions us to capture substantial long-term growth opportunities, transforming the usage of search and solidifying our leadership in the AI-driven search landscape.
Speaker Change: this strategic focus positions us to capture substantial long-term growth opportunities transforming the usage of search and solid buying our leadership in the ai driven search landscape
Robin Lee: Baidu core for our revenue growth slightly to R&B 26.7 billion in Q2.
Robin Li: Our one-stop shop for document creation is another example of how Baidu scales AI to, address real-world problems and generate substantial value.
Speaker Change: our one stop she for document quitation is another example of how bydo scales a i to address realworld problems and generate substantial valley
Robin Li: As our most pioneering internal product to embrace generative AI and LLMs, Baidu Wenku, is now re-emptying the benefits of its product renovation efforts.
Robin Lee: Thanks to the continuous acceleration of our AI cloud service, despite the headlines in our online margin events.
Robin Lee: We find that the most tangible benefit from AI comes from the adoption and use of applications built on top of LLS. Just three months ago, we asked that Ernie handled about 200 million API cost daily. Recently, it surpassed 600 million, or over 1 trillion tokens are generated every day. To make Ernie more affordable, we continue to expand our model portfolio and enhance our model capabilities to meet diverse customer needs. Last quarter, we launched three lightweight Ernie models, and they have quickly gained traction among enterprises and developers.
Speaker Change: as our most pioneering internal product to increase generated ai and lam by duen c is now reenting the benefits of its product renovation efforts
Robin Li: In the second quarter, Wenku's subscription revenue marked a year-over-year increase in, high teams.
Speaker Change: in the second quarter when whose subscription revenue marked a year-over year increase in highings
Robin Li: AI is rejuvenating the platform with comprehensive content understanding and generation capabilities, while continuously improving in versatility.
Speaker Change: ai is recruvenating the platform this comprehensive content understanding and generation capabilities
Robin Li: Since May, Baidu Wenku has launched new features resonating particularly well with young users, such as crafting lengthy documents with tens of thousands of characters, building and editing, PowerPoint presentations, and creating children's picture books with natural language guidance.
Speaker Change: while continuously improving a versatility
Speaker Change: since may by the whenco has launched new features resonating particularly well with young users
Speaker Change: such as crafting lensy documents with tens of thousands of characters building and editing powerpoint presentations and creating children's picture bos with natural language guidance
Robin Lee: Building on this momentum, we introduced Ernie 4.0 Turbo in June. It offers superior capabilities compared to Ernie 4.0 for typical use cases. Yet it is designed to be much cheaper and faster to run. Our current lineup now includes our flagship models, Ernie 3.5 and Ernie 4.0 and the enhanced model Ernie 4.0 Turbo and several lightweight models. This diverse portfolio allows us to accommodate the varying needs of our customers, optimizing for performance, cost, and latency.
Robin Li: The heightened engagement underscores Wenku's appeal to the next generation and signals, promising future growth.
Speaker Change: the high light
Speaker Change: the heightened engagement underscores when coth appeal to the next de generation and signals promising future growth
Robin Li: Moving from the digital realm to addressing changes and scaling AI in the physical world, our decade-long dedication to autonomous driving innovation and long-standing investment is bearing fruit.
Speaker Change: moving from the digital realm to addressing changes and sskinning ai in real in physical world our decadead l dedication to autonom striving innovation and last-standing investment
Robin Li: A polo goal, our autonomous ride hailing service, has recently achieved two significant breakthroughs.
Speaker Change: is barren foult
Robin Lee: In May, we made a strategic move to make API cost free of charge for three lightweight Ernie models. Ernie speed, Ernie light, and Ernie timing. In July, we significantly lowered the price of API cost for the two Ernie flagship models, Ernie 3.5 and Ernie 4.0. This decision is rooted in our commitment to enabling wider access to Ernie and making AI accessible to all. At the same time, we continue to lower the cost of model inference.
Speaker Change: a pollogo our autonomum slide heilling servies has recently achieved two significant breakthroughs
Robin Li: First, building a robust foundation for making commuting more affordable and benefiting more, people through technological innovation.
Speaker Change: if establishing a robust foundation for making commuting more affordable and benefiting more people through technological innovations
Robin Li: First, since June 19, building on our proven track record of safe operation, a polo goal, has successfully transitioned to offering 100% fully driverless ride hailing services in practically the entire Wuhan municipality. This means all vehicles are now operating without the need for human safety officers, on board, a major step forward in making autonomous ride-hailing spins commercially viable.
Speaker Change: on'm first
Speaker Change: things june nineteenths building our proven trat record of safe operation a pollogo has successfully transition to offering one hundred percent forty driver right heading services in practvely they entire han manysay palting
Robin Lee: One optimization of note, this quarter is the upgrade of pedal pedal. Our open source is learning framework to version 3.0. This upgrade significantly improves the framework's compatibility with our AI infrastructure and Ernie, which we expect will further help reduce model inference costs in the future. I briefly mentioned earlier that by doing skills, AI to address real-world problems and generate substantial value, both externally and internally. I like to discuss this in more detail.
Speaker Change: this means all vehicles are now operating theou the need for human safety officers on board
Speaker Change: a major staf forward in making autonom slide ha bins commercialally viable
Robin Li: Another milestone was the large-scale open-road testing of our sixth-generation autonomous, vehicle, the RT6.
Speaker Change: another milestone was the large scale open your testing of our six generation autonom to ve atle the partt six
Robin Li: Equipped with a battery swapping solution, RT6 is competitively priced at below 30,000 U.S. dollars for mass production. After thorough testing, we plan to officially roll out RT6 into our fleet, establishing a strong foundation for further substantial cost reductions in Apollo Go operations.
Robin Lee: Even with the ongoing investment in AI, Baidu core's non-gap operating margin improved by close to two percentage points year-over-year to 26% and non-gap operating profit grow by 8% year-over-year. Thanks to the operational efficiency gains, we continue to implement.
Speaker Change: equipped with a badway swatattkingum solution arguesix is competitively priced at below thirty thousand u s dollars formass production
Robin Lee: Despite the near-term pressure, we remain fully competent in our strategic direction and the transformation, the transformating potential of Gen AI and foundation models.
Speaker Change: after several testing we plan to officially roll out art six into our feet is establishing a strong foundation for further substantial cost reductions in our pollogo operations
Robin Lee: As we move further into 2024, we are happy to observe a significant change as Baidu is scaling AI to address real-world problems and generate substantial value both externally and internally.
Robin Lee: Externally, we have empowered our AI cloud customers to achieve greater efficiency and skillability by using Ernie. Our solutions have enabled clients to optimize their operations and realize significant benefits across various initiatives. Let me offer you some use cases to illustrate the real benefits problems we have addressed and how we add value to our customers. In the healthcare industry, Dr. Burmout is one of the greatest challenges. By leveraging early speed and our model builder, we help a healthcare automation solution provider to train and fine tune an industry-specific model that enables automatically generated medical records for doctors.
Robin Lee: Thanks to this effort, it's our commitment to making early increasingly affordable and accessible.
Robin Li: All in all, our efforts to scale AI to solve real-world problems fully align with our broader vision of creating a more efficient, equitable, and sustainable future.
Speaker Change: oimall our efforts to scale ai to solve real world problems fu fully align with our brother vision of quating more efficient equitable and sustainable future
Robin Lee: I'd like to highlight some key points on Baidu's The scaling of AI is accelerating at a breakneck pace, reflecting the real value that Ernie creates for people and for bins alike.
Robin Lee: We find that the most tangible benefit from AI comes from the adoption and use of applications built on top of LLS.
Speaker Change: looking at how our efforts yielding promising results we believe we are well positioned to capitalize on the growwing demand for ai-driven applications
Robin Li: Looking at how our efforts are yielding promising results, we believe we are well-positioned to capitalize on the growing demand for AI-driven applications.
Robin Lee: Just three months ago, we asked that Ernie handled about 200 million API cost daily. Recently, it surpassed 600 million, or over 1 trillion tokens are generated every day.
Robin Li: By leveraging AI to address complex challenges, we can continue our journey of innovation, and growth, and create long-lasting value for our stakeholders.
Speaker Change: by leveraging ai to address complex challenges we can continue our journey of innovation and growth and quate longlasting value for our stakeholders
Robin Li: I want to thank our employees for their effort and contribution, and our customers and shareholders for their trust in us.
Robin Lee: This solution significantly reduces the administrative burden on doctors, a major factor in burnout, and enhances clinical efficiency. After two months of model deployment, doctors using it were able to treat 50% more patients on average. In the recruitment industry, we have collaborated with the recruiting service company. By leveraging early capabilities for API costs, our customer has significantly upgraded the matching process between job descriptions and resumes. The smart matching upgrade has reduced the labor cost in this process by over 50%.
Speaker Change: i want to thank our employees for the effort and contribution and our customers and shareholders for their trustking us
Robin Li: Now, let me react to the key highlights for each business for the second quarter.
Speaker Change: now let me we ask the key highlights for each business for the second quarter
Robin Li: AI cloud revenue reached R&D 5.1 billion, marking a consecutive acceleration to 14% year-over-year growth, while sustaining non-gap operating profitability in the second quarter. The strong growth is mostly attributable to the following two factors. First, Gen-AI-related revenue continued its robust momentum, accounting for nearly 9% of our total AI cloud revenue in Q2, up from 6.9% in the previous quarter.
Speaker Change: ai cloud revenue reached rb five point one billion marking a consecutive acceleration to fourteen percent yoveryear growth while sustaining non-gaap operating profitability in the second quarter
Speaker Change: the strong growth is mostly a triotball to the followalling two factor
Speaker Change: first genera related revenue continued its robust momentum accounting for nearly n sense of our total ai cloud revenue in q two up from six point n percent in the previous quarter
Robin Lee: It is measured by total working hours while maintaining high quality results. This automation allows the customer to expand its services to a broader client base, thereby increasing revenue generating capability, and demonstrating the tangible benefits of integrating earnings into its operations. The customer is very satisfied with the result and is now exploring new collaboration opportunities. In the public service sector, we utilize the early speed to deliver tailored, personalized, and scalable services, addressing challenges like limited resources and efficiency needs.
Robin Li: As more enterprises integrate Gen-AI and foundation models into their daily operations, they increasingly come to us, thanks to our reputation as China's most advanced and cost-effective AI infrastructure provider, and our excellent math platform for model training and inference. During the quarter, we further advanced our AI infrastructure management, enhancing our ability to combine GPUs from more vendors, for optimal training and hosting of models. This ensures flexibility, reliability, and efficiency, positioning us to capture a larger share of the GPU cloud and LLM market.
Speaker Change: as more enterprises inweight geni and foundundation models into their daily operations
Speaker Change: they increase me come to us thank to our reputation as china's most advanced and cost effective ai infrastructure provider and our excellent maass platform for model twinaining and inference
Speaker Change: seven
Speaker Change: during the quarter we further advanced our ai infrastructure management
Speaker Change: in intensing our ability to combine gp u from more vendor for optimal training and hosting of models
Speaker Change: this inshres flexibility reliability and efficiency positioning us to capture larger share of the gpo cloud and lm market
Robin Lee: This is particularly valuable in China's grassroots public services, where shortages in manpower make large-scale personalized services challenging. We have collaborated with the customer to help over 6,000 villages in China improve a public service, offering citizens more personalized and efficient support. Since the large-scale launch of the service in April, its daily usage has surged over 15 fold to more than 2 million times per day. Helping locals handle tasks such as household registrations, social security inquiries, and tax declarations.
Speaker Change: we also continued to develop the toolcates on our mass platform for our customers and partners enabling the earnning family of models to deliver a furea priide performance ratio
Robin Li: We also continued to develop toolkits on our math platform for our customers and partners, partners, enabling the early family of models to deliver a superior price-performance ratio.
Robin Li: We're proud that our model builder now supports the full model development lifecycle, including, data management, fine-tuning, evaluation, optimization, and prompt engineering.
Speaker Change: we proud that our modelbuildder now supports the full model development lifecycle including data management fing to me evaluation optimization and prompt engineer
Robin Li: A major upgrade has been the introduction of diverse hybrid training data science, spanning from general to specialized industry-tailored data sets. This enables efficient fine-tuning and ensures high performance for industry-specific applications, while retaining strong general LLM capabilities.
Speaker Change: a major upweade as being the introduction of dibirth high brade training data size
Speaker Change: tspending from jal to specialized industry tailored dataasfence
Robin Lee: To make Ernie more affordable, we continue to expand our model portfolio and enhance our model capabilities to meet diverse customer needs. Last quarter, we launched three lightweight Ernie models, and they have quickly gained traction among enterprises and developers. Building on this momentum, we introduced Ernie 4.0 Turbo in June. It offers superior capabilities compared to Ernie 4.0 for typical use cases. Yet it is designed to be much cheaper and faster to run.
Robin Lee: Internally, we have accelerated the renovation of title search with Ernie, significantly enhancing our consumer-facing products at large scale. Generative AI is transforming user search experience, pushing beyond traditional boundaries. Since the second half of last year, we have been testing early-powered features on a small scale. This quarter are focused as shifted to substantially amplifying search capabilities with advanced AI features. Currently, 18% of search without pages contain generated content. Up from 11% in mid-May.
Speaker Change: this enables efficient find to me and ensureres high performance for industriryes specific applications while retaining strong journal and capabilities
Robin Lee: Our current lineup now includes our flagship models, Ernie 3.5 and Ernie 4.0 and the enhanced model Ernie 4.0 Turbo and several lightweight models. This diverse portfolio allows us to accommodate the varying needs of our customers, optimizing for performance, cost, and latency.
Robin Li: Also, thanks to our continuous refinement to App Builder, the number of AI-native apps, on our cloud now runs into the hundreds of thousands. These applications span a wide array of industries and scenarios, from online education and e-commerce, to public service sectors, and are integrated into both online platforms and smart devices.
Robin Lee: In May, we made a strategic move to make API cost free of charge for three lightweight Ernie models. Ernie speed, Ernie light, and Ernie timing.
Speaker Change: also sex to our continuous refinement to abbuer the number of a i native ad are expoued now runing to the hundreds of thousands
Robin Lee: In July, we significantly lowered the price of API cost for the two Ernie flagship models, Ernie 3.5 and Ernie 4.0. This decision is rooted in our commitment to enabling wider access to Ernie and making AI accessible to all.
Speaker Change: list applications spend out white ray of industries and scenaralss from online education and e-commerce to public service sectors
Robin Lee: At the same time, we continue to lower the cost of model inference.
Speaker Change: and are integrated into both online platforms and smart devices
Robin Lee: One optimization of note, this quarter is the upgrade of pedal pedal.
Robin Li: The second major driver of AI Cloud revenue acceleration is cross-selling of our CPU Cloud, services to our GPU Cloud customers. In Q2, we continue to observe an increase in CPU spending among our GPU Cloud customers.
Robin Lee: This AI generated search without delivering more accurate and direct answers, enhancing content quality and providing previously unattainable information. This improvement has led to increased user satisfaction and engagement. As more users are now turning to Baidu Search for more complex queries, increasing Baidu Search versatility. AI generated search without may reduce app impressions and therefore have a negative short-term impact on monetization, but they provide significant value to our users. By prioritizing long-term user experience over immediate revenue and profits, we see the strategic implementation of generated search as essential to driving future success.
Speaker Change: the second major driver of ai cloud revenue acceleration is cross-selling up our cpu cloud services to our gp cloud customers
Speaker Change: in q two we continue to observe an increase in cpuo spending among our gpu cloud customers
Robin Li: We see that our strong brand recognition in GPU Cloud is helping us win businesses in, the CPU Cloud industry.
Speaker Change: we see that our strong brand regnition in gpuo cloud is helping us wining busensus in the cpu cloud industry
Robin Li: With revenue acceleration, AI Cloud businesses continue to deliver positive non-gap operating, profit and improved margins.
Speaker Change: with revenue acceleration aicl business continue to deliver positive non-gaap operating profit and improved margins
Robin Li: Legacy Cloud saw margin expansion, and gen-AI-related businesses are expected to have a higher normalized, margin compared to the traditional cloud service.
Speaker Change: leg acsay cloud saw margin expansion and gena i related bbuses expected to have higher normalized the margin compared to the traditional cloud service
Robin Li: Overall, we remain confident in the strong growth outlook for our AI Cloud revenue, and, we aim to continue generating non-gap operating profit going forward.
Speaker Change: overall we remain confident ain the strong growth outlook for our aaccl revenue and we aim to continue generating non-gaap operating profit going forward
Robin Lee: Our AI investments are fostering deeper user interactions. New interactive features enable users to refine their questions through multi-round conversations and enhancing the overall user experience. Additionally, there is a significant increase in users utilizing text and image creation tools within search. This development opened doors for long-term value creation. Last quarter, I mentioned that we planned to accelerate the distribution of earning agents. So far, we have seen more and more developers, advertisers and partners, together on our platform to develop innovative AI agents.
Robin Li: For our mobile ecosystem, FIDO Core's online marketing revenue declined by 2% year-over-year, in the second quarter due to broader macroeconomic challenges, competition, and our aggressive, AI-driven search renovation.
Speaker Change: for mobile ecosystem iwill course ponline marketing revenue declined by two percent year over year in the second quarter due to broader macoeconomic challenges competition and our aggressive ai driven search granimation
Robin Li: Key offline verticals such as real estate, franchising, and automobile remained subdued.
Speaker Change: key offline verticals such as real estate franchising and automobile remained subdued
Robin Li: Despite these challenges, we remain committed to optimizing our operations to maintain both, a healthy margin and cash flow while accelerating the AI-native transformation of our products.
Speaker Change: despite this challenges we remain committed to optimizing our operations to maintain both a healthy margin and cash flow while accelerating the ai maing transformation of our products
Robin Li: Incremental ad revenue from gen-AI and LLM enhancements to our advertising system continue, to grow quarter-over-quarter, driven by the ongoing reconstruction of our monetization system and marketing capabilities.
Speaker Change: in q two incremental ad revenue from gena i in an l l and enhancements to our advertising system continue to grow quarter over quarter driven by the ongoing reconstruction of our twenty five bastern system and marketing counforn
Robin Lee: The earning agent distribution within Baidu increased dramatically in July, exceeding 8 million daily, more than double the number in May. Currently, the most frequently used agents by users for content creation, personal insights such as personality testing and tools such as translation and schedule planning. At the same time, we also see a joint of earning agents being more widely developed and applied both internally and externally. Our first college application assistant agent is one example of how earning agents are being distributed within Baidu to assist users with complex problem solving and decision making.
Robin Li: We envision Earning Agent as a compelling opportunity for our ad business. Earning Agent has the potential to revolutionize our traditional CPC model into a more efficient, CPS model in the future, as it transitions from pre-sale consultations to catalyzing more direct sales on our platform. Serving as a virtual salesperson, Earning Agent empowers advertisers to provide more, personalized content and product introduction, thereby ensuring better pre-sale consultation.
Speaker Change: we nion earning agent as a compeing opportunity for adments
Speaker Change: earning agent as the potential to revolish n our traditional cpc model into a more efficient cps model in the future
Speaker Change: as transitions from presale consultations to catalying more direct sales are platform
Speaker Change: learning as a virtual sales of person orning agent inempowers advertisers to provide more personalized content and product introduction
Speaker Change: weby ensuring better preale consultation
Robin Li: Early adopters from the education, legal, and B2B sectors are pioneering Earning Agent, and have already seen a notable increase in effective sales leads.
Speaker Change: early adopters from the education legal and btwob sectors are pioneering earning agent and have already seen a notable increase in effective cell ceales lead
Robin Lee: At the year after China's National College entrance examination. University after checking exam results is crucial for the 10-million-plus candidates. In June, Baidu launched an innovative college application assistant agent, which is designed to meet personalized needs for selecting universities and majors. Following the National College entrance exam, the peak daily active users of the agent approached 2 million highlighting its significant impact and utility. While all these new features for more sophisticated user needs have not yet been monetized, they are transformative for us, for search, and a crucial for our future success.
Robin Li: We're proud to see that 16,000 advertisers already have their own Earning Agents, which, are now being distributed on the Baidu platform.
Speaker Change: we' proud to see that sixteen thousand advertisers already have their on orning agent
Speaker Change: which are now being distributed on the bible calform
Robin Li: Looking ahead, we believe Earning Agent is poised to deepen its penetration and broaden, its reach across more sectors, unlocking promising potential for our advertising business.
Speaker Change: looking ahead we believe earning agent if
Speaker Change: poised to deepen its penetration and broaden its breage across more effactor unlocking promising potential for our advertising
Robin Li: And moving into intelligent driving, I mentioned earlier that ApolloGo has made remarkable, breakthroughs in Wuhan and achieved 100% fully driverless operations in practically the entire municipality, reducing Q2 costs per vehicle by more than half compared to Q2 of last year.
Speaker Change: and moving into intelelig dryving i mentioned earlier that a pologo have made remark o breaksu
Speaker Change: inwill hind achieved one hundred hundred percent fully graphates operations in practically the enentirement municipality reducing q two costs to vehicle by more than half compared to q two of af years
Robin Li: On top of that, ApolloGo has started scalable testing of the latest RT6 vehicles on open, roads. These two factors are critical milestones in our strategy aimed at reaching UE breakeven, in targeted cities.
Robin Lee: This strategic focus positions us to capture substantial long-term growth opportunities, transforming the usage of search and solidifying our leadership in the AI-driven search landscape. Our one-stop shaft for document creation is another example of how Baidu scales AI to address real-world problems and generate substantial value. As our most pioneering internal product to embrace generating AI and LLM, Baidu Wenku is now re-amping the benefits of its product renovation efforts. In the second quarter, Wenku's subscription revenue marked a year over increase in high teams.
Speaker Change: a comof that a pollogo has started scalable testing of therelative to art six vehicles on open goals
Speaker Change: this two factors are critical milestones in our strategy aimed at whiching eue breaking them in targeted cities
Robin Li: In addition to cost reduction, ApolloGo's operations continue to offer more expedient, services.
Speaker Change: in addition to a cost reduction apollo goals' operations continue to offer more expedient services
Robin Li: In Wuhan, its services is now available to 9 million people.
Speaker Change: in oham exservited it's now available to nine million people
Robin Li: The number of ApolloGo pickup points at the end of June has increased by over threefold, from the previous quarter, enhancing accessibility and convenience for passengers.
Speaker Change: the number of apollogoust the artpoints at the end of je has increased by over three fold from the previous quarter insensing accessability and convenience folcpassinginterest
Robin Lee: AI is re-groupinating the platform with comprehensive content understanding and generation capabilities, while continuously improving in versatility. Since May, Baidu Wenku has launched new features resonating particularly well with young users. Such as crafting lengthy documents with tens of thousands of characters, building and editing PowerPoint presentations, and creating children's picture books with natural language guidance. The heightened engagement underscores Wenku's appeal to the next generation and signals promising future growth. Moving from the digital realm to addressing changes and scaling AI in physical world, our decade-long dedication to autonomous driving innovation and long-standing investment is bearing fruit.
Robin Li: The deepening of our operational footprint and increased station density have propelled, utilization rate for our vehicles, significantly driving up the daily rides per vehicle and the distance per ride.
Speaker Change: the deepening of our operational footprint and increased stian density have propelled utilization rate for our vehicles
Speaker Change: significantly driving up that they need rise per vehicle and the distance prorived
Robin Li: Nationwide, ApolloGo provided about 899,000 rides to the public in the second quarter, marking a 26% year-over-year increase.
Speaker Change: nationwide a pollog provided about eight hundred ninety nine bound rights to the public in the second quarter marking a twenty th six percent year over year increase in july the accumulated ridees provided to the public surpast seven minute
Robin Li: In July, the cumulative rides provided to the public surpassed 7 million.
Robin Li: Even with these milestones, our share in the entire ride-hailing service market is very, small. It will take many years for us to reach a meaningful market share in China or elsewhere.
Speaker Change: even withid this milestones our share in the entire right-heading service market is very small it will hit many years for us to reach a meaningful market share in china or elsewhere
Robin Li: Looking ahead, we are committed to providing increasingly affordable, convenient, and safe, travel for more passengers and drive long-term, sustainable growth.
Speaker Change: looking ahead we are committed to providing increasingly affordable convenient and safe travel for m passengers and drive long-term sustainable growth
Robin Li: We will remain resolute in executing on our operational strategy, aimed at boosting efficiency, and steering our intelligent driving bins towards profitability. In summary, by extrapolating the ongoing progress, from leveraging earnings to revolutionize, product usage, and transform bins operations for our users, customers, and developers, and driven by a deep belief in technological innovation, we aim to create value for society and contribute to the greater good.
Speaker Change: we will remain resolute in executing on our operational strategy aimed at oasting efficiency as ding our intelligion driving ben towards profitability
Robin Lee: Apollo Gaul, our autonomous ride hailing service, has recently achieved two significant breakthroughs. If establishing a robust foundation for making commuting more affordable and benefiting more people through technological innovations. First, things June 19. Building our proven track record of safe operation Apollo Gaul has successfully transitioned to offering 100% fully driverless ride hailing services in June, practically the entire Wuhan Munisi Palace. This means all vehicles are now operating without the need for a human safety officer on board, a major death forward in making autonomous ride hailing bins commercially viable.
Speaker Change: in summary by extrapolating the ongoing progress
Speaker Change: from leveraging earning to revolutioniz product usage and transform bins operations for our users customers and developers and driven by a deep belief in technological innovation
Speaker Change: we aim to create value for society and contribute to the quater good
Robin Li: With that, let me turn the call over to Rong to go through the financial results.
Speaker Change: with that let me turn the call over to them to go through the financial results
Rong Luo: Thank you, Robin.
Rong Luo: Now, let me walk through the details of our second quarter financial results. Total revenue was RMB33.9 billion, which was basically flat from last year. Revenue from Baidu Core was RMB26.7 billion, increasing 1% year-over-year. Baidu Core's online marketing revenue was RMB19.2 billion, decreasing 2% year-over-year. Baidu Core's non-online marketing revenue was RMB7.5 billion, up 10% year-over-year, mainly driven by the AI cloud business.
Them: thanker rob and analm we was through er details up our second quarter finanal results
Speaker Change: total revenue was ion the crepment billion which was basically fllex from last year our revenue from bital qu was twenty six and seven billion increasing one percent year over- year
Speaker Change: by co only marketing revenue was in thousandnine point twoper then decreasing to prevent year over year by who cocourse now only marking revenue was i bdesevenm point five billan up ten percent year over-year mainly driven by the air cliness
Robin Lee: Another milestone was the large scale open road testing of our six-generation autonomous vehicle, the RT-6. It could be a bad-waste watching solution, RT-6 is competitively priced at below $30,000 US dollars for mass production. After several testing, we plan to officially roll out RT-6 into our fleet, establishing a strong foundation for further substantial cost reduction in our political operations. All in all, our efforts to scale AI to solve real-world problems fully align with our broader vision of creating a more efficient, equitable, and sustainable future.
Rong Luo: Revenue from IT was RMB7.4 billion, decreasing 5% year-over-year.
Speaker Change: around the fivesee was irb sevenm four and decreasing five percent year over-year caal revenue was i on b sixteen point four bill increasing one percent year over-year primary due to an increase in traffic acquisition causes a caes related to a cloud business
Rong Luo: Cost of revenue was RMB16.4 billion, increasing 1% year-over-year, primarily due to an increase, in traffic acquisition costs and the costs related to AI cloud business.
Rong Luo: Operating expenses were RMB11.6 billion, decreasing 9% year-over-year, primarily due to a decrease, in personnel-related expenses, expected credit losses, and the channel spending and promotional marketing expenses. Baidu Core's operating expenses were RMB10.2 billion, decreasing 10% year-over-year. Baidu Core's SG&A expenses were RMB4.8 billion, decreasing 11% year-over-year.
Speaker Change: operating expenses where ion b eleven point six billion decreasingin nine percent year overyear primary due to a decrease in personnel relativeating expenses exped cred loice and the channel spending and promotional marketing expenses
Speaker Change: by the cost operating expenses were on the ten point point two bill decreas in ten percent year over-year
Robin Lee: Looking at how our efforts are yielding promising results, we believe we are well positioned to capitalize on the growing demand for AI-driven applications. By leveraging AI to address complex challenges, we can continue our journey of innovation and growth and create long-lasting value for our stakeholders.
Speaker Change: by the t asri expenses
Speaker Change: why before p april and decreasing eleven percent year over-year sjunior counting for eighteen percent one a up by those cost revenue engl this quarter comppeed with twenty percent in the same yearer last year
Rong Luo: SG&A accounted for 18% 1A of Baidu Core's revenue in this quarter, compared to 20% in, the same period last year.
Rong Luo: Baidu Core's R&D expenses were RMB5.4 billion, decreasing 8% year-over-year. R&D accounted for 20% of Baidu Core's revenue in the quarter, compared to 23% in the same, period last year.
Speaker Change: by the call ired expenses why i m b five point four billion decreasing eight percent year over-year i the accounting for twenty percent of b ble calls through revenue in the quarter compared to twenty three percent this im period last year
Robin Lee: I want to thank our employees for their efforts and contributions and our customers and shareholders for their trust in us.
Rong Luo: Operating income was RMB5.9 billion, Baidu Core's operating income was RMB5.6 billion, and Baidu Core's operating margin was 21%. Non-GAAP operating income was RMB7.5 billion, Non-GAAP Baidu Core's operating income was, RMB7 billion, and Non-GAAP Baidu Core's operating margin was 26%.
Speaker Change: operating income was ib ied point nine b by the cost operating income was i bide twenty six billion and by the cause operating margin was twenty-one percent
Robin Lee: Now, let me re-ask the key highlights for each business for the second quarter. AI cloud revenue reached RB 5.1 billion, marking a consecutive acceleration to 14% over your growth while sustaining non-gab operating profitability in the second quarter. The strong growth is mostly attributable to the following two factors. First, Gen-AI-related revenue continued its robust momentum, accounting for nearly 9% of our total AI cloud revenue in Q2, up from 6.9% in the previous quarter.
Speaker Change: nonga oper income who was on desome point five better non thereby the co operating income was in seven billion and non them by the co operating margin was twenty- six percent
Rong Luo: Total other income net was RMB771 million, decreasing 44% year-over-year, primarily due, to a decrease in net foreign exchange gain and disposal gain, partially offset by a decrease in fair value loss and impairment loss from long-term investments.
Speaker Change: total other income net was on b seven hundred seventy one million decreasing forty four percent year overyear primaryily due to a decrease in net for foreign exchange gain and is pos again
Speaker Change: partially offset by decreasing fair value loss and impairment loss from long-term investments
Rong Luo: Income tax expenses was RMB1.1 billion, compared to RMB1.3 billion in the same period last, year.
Speaker Change: income tax expenses was rb one point one billion compared to only one point three p in a same period last year
Rong Luo: Net income attributable to Baidu was RMB5.5 billion, and diluting earnings per ADS was, RMB15.01.
Speaker Change: net income attributable to buy to was i the five point five billion and diluting earnings pre years were on be fifteen point zero one
Robin Lee: Our open source is learning framework to version 3.0. This upgrade significantly improves the framework's compatibility with our AI infrastructure and Ernie, which we expect will further help reduce model inference costs in the future.
Robin Lee: As more enterprises ink with Gen-AI and Foundation models into their daily operations, they increasingly come to us thanks to our reputation as China's most advanced and cost effective AI infrastructure providers and our excellent math platform for model training and inference. During the quarter, we further advanced our AI infrastructure management, intencing our ability to combine GPUs from more vendors for optimal training and hosting of models. This ensures flexibility, reliability and efficiency, positioning us to capture a larger share of the GPU cloud and LLM market.
Rong Luo: Net income attributable to Baidu Core was RMB5.5 billion, and net margin for Baidu Core was, Longgan Net Income attributable to Baidu was RMB 7.4 billion, Longgan Diluting Earnings per ADS was RMB 21.02, Longgan Net Income attributable to Baidu Corp was RMB 7.3 billion, and Longgan Net Margin for Baidu Corp was 27%.
Speaker Change: net income attributable to b the call was i m b fiveineppointing five billten a net moin while by thequ was twenty percent
Speaker Change: in on get net income attributable to buy was high the seven point four billion nongaap diluting earnings p years were on twenty one point zero two
Robin Lee: I briefly mentioned earlier that by doing skills, AI to address real-world problems and generate substantial value, both externally and internally.
Speaker Change: now again that income achieved the able to bby to core was i m b seven point three bllan and nonga net margin for byoqu was twenty- seven percent
Rong Luo: As of June 30, 2024, cash, cash equivalents, restricted cash, and shorting investments were RMB 162 billion, and cash, cash equivalents, restricted cash, and shorting investments excluding IT were RMB 155 billion.
Speaker Change: us ofth june thires twenty andtwenty four cash cash equivalence withry cash and shoting investments were imed the one hundred and sixty two billion and cash cash equivalence
Robin Lee: I like to discuss this in more detail.
Speaker Change: restry catchash a shopping investments it cruly it were i on the one hundred fifty five but
Rong Luo: Free cash floor was RMB 6.3 billion, and free cash floor excluding IT was RMB 5.9 billion.
Robin Lee: We also continued to develop the toolkit on our math platform for our customers and partners. Innebling the early family of models to deliver a superior price performance ratio. We're proud that our model builder now supports the full model development lifecycle, including data management, fine tuning, evaluation, optimization, and prompt engineering. A major upgrade has been the introduction of diverse high-grade training data sites, spanning from general to specialized industry tailored data sets. This enables efficient fine tuning and ensures high performance for industry specific applications while retaining strong general and capabilities.
Speaker Change: free cash flowor was iring the six p three bill and free cash flow cluding it was onb five planma b
Rong Luo: Finally, Baidu Corp had approximately 31,000 employees as of June 30, 2024.
Speaker Change: finally by two call approimmeury certainty one solll employees as south june thir twentthousand twenty four westst operators less now open a court to questents
Operator: With that, operators, let's now open the call to questions.
Operator: Ladies and gentlemen, we will now begin the question and answer session.
Speaker Change: ladies and gentlemen we will now begin the question ancessation
Operator: If you wish to ask a question, you will need to press star 1 on your telephone and wait for, your name to be announced.
Speaker Change: if you wish to ask a question you ll need to pase starle one on your telephone and waitforyourname to be announced
Operator: If you wish to cancel your request, please press star 2.
Speaker Change: if you wish to cancel your request please ase start to
Operator: Your first question comes from Alex Yao with JPMorgan.
Speaker Change: your first question comes from alex yiao with jp morgan please ask your question
Alex Yao: Please ask your question.
alex yiao: thank you ment for taking a question i have a view on a year development so first of all how do you guys foresee the competitioning ai evolving over the next two to three igyears
Alex Yao: Thank you, Manjuan, for taking my question.
Alex Yao: I have a few on AI development.
Robin Lee: Also, thanks to our continuous refinement to app builder, the number of AI native apps on our cloud now runs into the hundreds of thousands. These applications span a wide array of industries and scenarios, from online education and e-commerce to public service sectors, and are integrated into both online platforms and smart devices. The second major driver of AI cloud revenue acceleration is cross-selling of our CPU cloud services to our GPU cloud customers.
Speaker Change: we buite those occurrent first mover advantagees sustthen or have the technology gap among key players narrow over time
Speaker Change: how do you guys view the current compeit landscape computer
Speaker Change: to horizontal kimi to baba and tensson in addition to technology leadership what other factors are crucial topute a success or ecosystem in the industry thank you
Alex Yao: So first of all, how do you guys foresee the competition in AI evolving over the next two to three years?
Speaker Change: hi anexis is rabin competition well be fears over the next to years to do three years other mid i think ever can make money from it while webe able to salive
Robin Lee: In Q2, we continue to observe an increase in CPU spending among our GPU cloud customers. We see that our strong brand recognition in GPU cloud is helping us win businesses in the CPU cloud industry. With brand new acceleration, AI cloud business continue to deliver positive non-gap operating profit and improved margins. Legacy cloud saw margin expansion and gen AI-related businesses are expected to have higher normalized margin compared to the traditional cloud service. Overall, we remain confident in the strong growth outlook for our AI cloud revenue and we aim to continue generating non-gap operating profit going forward.
Alex Yao: Will Baidu's current first mover advantage sustain, or will the technology gap, among key players narrow over time?
Speaker Change: last year when china's ai industry was still at the so called waroff and with foundation modells period i predicted that the industry would go through consolidation and only a few foundation modells would survive
Alex Yao: How do you guys view the current competitive landscape compared to, for example, Kimi, Doubao, BABA, and Tencent?
Speaker Change: we are seeing strong demand for our generning i products and i'm proud that or we have continued to strengthening and market lead
Speaker Change: and i'm confident we while maintaining our leadership going forward
Speaker Change: our decade long investment in ai provided us with the first mo advantage and tip technological leadership
Robin Lee: For mobile ecosystem, title course online marketing revenue declined by 2% year-over-year in the second quarter due to broader macroeconomic challenges, competition, and our aggressive AI driven search renovation. Key offline verticals such as real estate, franchising, and automobile remained subdued. Despite these challenges, we remain committed to optimizing our operations to maintain both a healthy margin and cash flow while accelerating the AI-nating transformation of our products. In Q2, incremental ad revenue from gen AI and LLM enhancements to our advertising system continue to grow quarter-over-quarter driven by the ongoing reconstruction of our monetization system and marketing cloud.
Alex Yao: In addition to technology leadership, what other factors are crucial to build a successful ecosystem in the AI industry?
Speaker Change: we were the first public company globally to launch a gt like model in march of last year
Alex Yao: Thank you.
Speaker Change: and things thing thing we kept leveraging our self about the four layer i architecture for a model upgrade
Robin Li: Hi, Alex.
Speaker Change: last october we launched a earning for ondarro that's china's first iv four type model and this june we
Robin Li: This is Robin.
Robin Li: Competition will be fierce over the next two to three years.
Speaker Change: plunch the earning four one zero turable which offers superior capabilityities over for onbureau
Robin Li: Ultimately, I think whoever can make money from it will be able to survive.
Speaker Change: we will continue to grow out adethquakated depersiions of our fllexing foundation models
Speaker Change: and our application driven approach further expand our competitive advantage and set us apart from the others
Robin Lee: Externally, we have empowered our AI cloud customers to achieve greater efficiency and skillability by using Ernie.
Robin Lee: We, Ignatian, Ernie Agent, as a compelling opportunity for our admins. Ernie Agent, as a potential to revolutionize our traditional CPC model into a more efficient CPS model in the future. As it transitions from pre-sale consultations to catalyzing more direct cells, our platform. Serving as a virtual cells person, Ernie Agent empowers advertisers to provide more personalized content and product introduction, thereby ensuring better pre-sale consultations. Early adopters from the education, legal and B2B sectors are pioneering Ernie Agent and have already seen a notable increase in effective cell fleets.
Speaker Change: we believe that foundation models hold little practical value without real applications be would a upon land
Robin Li: Last year, when China's AI industry was still at the so-called war of 100 foundation models period, I predicted that the industry would go through consolidation, and only a few foundation models would survive.
Robin Lee: Our solutions have enabled clients to optimize their operations and realize significant benefits across various initiatives.
Speaker Change: widely used by users and customers
Speaker Change: so i mention earlier
Speaker Change: that currently earie hans about six hundred million api costs and generates about one trillion tokens every day which we believe is the highest in china
Robin Lee: Let me offer you some use cases to illustrate the real benefits problems we have addressed and how we add value to our customers.
Speaker Change: we have being heavily using earning to renovate our own products like by do search or by will and cool into a i native applications a large scale
Speaker Change: and this approach has helped up gain valuable insights from users and
Speaker Change: continuously supportedve as the upgrade up earning and this will experience allows us to create betterther solutions for our external customers when they you lear
Robin Lee: We're proud to see that 16,000 advertisers already have their own Ernie Agents, which are now being distributed on the Baidu platform. Looking ahead, we believe Ernie Agent is poised to deepen its penetration and broaden its reach across more sectors, unlocking promising potential for our advertising benefits.
Robin Li: We are seeing a strong demand for our gen AI products, and I'm proud that Ernie has continued to strengthen its market lead, and I'm confident we will maintain our leadership going forward. Our decade-long investment in AI provided us with the first mover advantage and technological leadership.
Speaker Change: at the same time we have being working really hard to make earning increasingly accessible and affordable so it can be used widely to address real world needs
Speaker Change: we continue to lower the cost of modern infrs expand our mono california and
Robin Lee: And moving into intelligent driving, I mentioned earlier that Apollo Go has made remarkable breakthroughs in Wuhan, achieved 100% fully-dramanized operations in practically the entire municipality, reducing Q2 cost from vehicle by more than half compared to Q2 of last year. On top of that, Apollo Go has started scalable testing of the latest RG6 vehicles on open roads. These two factors are critical milestones in our strategy aimed at reaching UE breakeven in targeted cities.
Speaker Change: develop two ks for a model building and atbuilding so that our customers partners developers can leverage to the power of ning more efficiently and effectively
Speaker Change: and this application driven approach has how that create a dynamic and expanding ecosystem
Speaker Change: by the end of june pedal pedo and earnie developedper community had brwn to around the fourteen fourteen point seven minute
Speaker Change: meanwhile handles of thousands of enterprises are now engaged with us highlighting the expanding inincls and utility of our ecosystem
Robin Lee: In addition to cost reduction, Apollo Go's operations continue to offer more expedient services. In Wuhan, its services is now available to 9 million people. The number of Apollo Go's pick-up points at the end of June has increased by over threefold from the previous quarter, enhancing accessibility and convenience for passengers. The deepening of our operational footprint and increased station density have propelled utilization rate for our vehicles, significantly driving up the daily rides per vehicle and the distance per ride.
Speaker Change: looking ahead we will be patient we will continue to iterate on earnings capabilities to ensure we stay as a forefront of this new ai wave
Speaker Change: by leveraging realworld applications and feedbackgrfrom marecosystem partners
Speaker Change: gain valuable resources and insights to continuously refine our model
Speaker Change: the synerity among our wbrantand ecosystem
Speaker Change: advan the model capabilities and an application driven approach
Robin Lee: Nationwide, Apollo Go provided about 899,000 rides to the public in the second quarter, marking a 26% year-over-year increase. In July, the cumulative rides provided to the public surpassed 7 million. Even with these milestones, our share in the entire ride-hating service market is very small. It will take many years for us to reach a meaningful market share in China or elsewhere. Looking ahead, we are committed to providing increasingly affordable, convenient and safe travel for more passengers and drive long-term, sustainable growth.
Speaker Change: well continueed to strengthen then our leadership in this rapputly in this rabbitidly evolving ai landscape
Speaker Change: thank you
gary you: next question comes from gary you with morgan stanley please ask your questions
gary you: hi thank you for opportunity to ask questions on my questions are related to ei and search
Gary Yu: so what kind of progress has been made in renovating traditional search using jai and the last language models what kind of kbi is the company in mononeittoring for ei in search and have you haveobserved any changes in useer behaviors you implementing i
Robin Lee: We will remain resolute in executing on our operational strategy, and at hosting efficiency and steering our intelligent driving bins towards profitability. In summary, by extrapolating the ongoing progress, from leveraging earnings to revolutionize product usage, and transform bins operations for our users, customers, and developers, and driven by a deep belief in technological innovation, we aim to create value for society and contribute to the greater good.
Gary Yu: in the search our business
Gary Yu: related to that what is the lower plane for i search moving forward and it s specific ki that the company aim to achieve at the end of the year for example certain percentage of search we so generated by
Speaker Change: yeah we are very committed to using cutting edgeof the technologies to transform our search capabilities and blend searchly speed in one that so kv eyes related to buy to adp well overrider search only tpvi
Robin Lee: Let me turn the call over to Rong to go through the financial results.
Rong Luo: Thank you, Robin. [inaudible] Nonggai Net income, attributable to Baidu was RMB 7.4 billion. Nonggai diluting earnings for ADS was RMB 21.02.
Speaker Change: in the second quarter we accelerated the pace off renovating our search by leveraging earning our primary goal if to improve user experience offer more capabilities and boost user engagement by doing
Robin Li: You know, we were the first public company globally to launch a GPT-like model in March of, last year.
Robin Li: Since then, we kept leveraging our self-developed four-layer AI architecture for model upgrade. Last October, we launched Ernie 4.0.
Robin Li: That's China's first GPT-4 type model.
Speaker Change: as highlighted in my opening remarks we've made several important in progresses
Robin Li: And this June, we launched Ernie 4.0 Turbo, which offers superior capabilities over 4.0.
Speaker Change: first we we are providing smarter search results at an expanding pace
Robin Li: We will continue to roll out upgraded versions of our flagship foundation models.
Speaker Change: currently eighte percent off search without i provided by genuallyi
Robin Li: And our application driven approach further expands our competitive advantage and sets us apart from the others.
Speaker Change: this without our more competensive and relevant than traditional search withoutts
Robin Li: We believe that foundation models hold little practical value without real applications built upon them that are widely used by users and customers.
Speaker Change: offering information that was previously unavailable or scattered around different places
Speaker Change: furthermore generai enables infinite supply of conant ondemand
Speaker Change: and they can effectively catch the decline of
Speaker Change: the openweb system
Speaker Change: second when gener results can already meet user needs from the curconquiry we will recommend other contents or services just like a user would find in a typical new feed the context
Speaker Change: this while enhan users time spend and retention
Speaker Change: third the new interactive function in by d search allows users to refine their questions and follow off smaam conversations
Speaker Change: reshaping how users express their need and making the search experience more natural and efficient
Robin Li: So I mentioned earlier that currently Ernie handles about 600 million API calls and generates about 1 trillion tokens every day, which we believe is the highest in China.
Speaker Change: but
Speaker Change: additionally i'would like to point out that we are rapidly rolling out earning agents with in search without to provide intelligent assistance to our users
Robin Li: We have been heavily using Ernie to renovate our own products like Baidu Search or Baidu Lengku into AI native applications on a large scale. And this approach has helped us gain valuable insights from users and continuously supported, the upgrade of Ernie.
Robin Li: And this experience allows us to create better solutions for our external customers when they use Ernie.
Speaker Change: by opening up our platform devidper can create a wide range of agents payilored to meet the unique need of each and every user
Robin Li: At the same time, we have been working really hard to make Ernie increasingly accessible and affordable so it can be used widely to address real world needs. We continue to lower the cost of model inference, expand our model portfolio, and develop toolkits, for model building and app building so that our customers, partners, developers can leverage the power of Ernie more efficiently and effectively.
Speaker Change: our vision is to create an ecosystem where many dedevelopers contribute to building a robust network of agents
Speaker Change: collectively in reaching the user experience and outperform the most powerful fronttier models
Speaker Change: addressing a broad spectrum of challenges and opportunity
Speaker Change: this all the effort mentioned the types of search clers bu have become more diversified
Speaker Change: users are inquasly posing complex questions
Speaker Change: seeking comprehensive and invest answers from bio search
Speaker Change: while this demands were once challenging to meet they can now be efficiently addressed
Rong Luo: Nonggai Net income, attributable to Baidu core was RMB 7.3 billion, and Nonggai Net margin for Baidu core was 27%. As of June 30, 2024, cash equivalence was[inaudible] with our pre-test, let's not open the clock to questions.
Speaker Change: on how about that users are also using by due to create a content
Speaker Change: such as fiding and editing text and images general eyes is transforming by research into a multiffunctional creative topform offering a wider ranch
Robin Li: And this application driven approach has helped us create a dynamic and expanding ecosystem.
Robin Li: By the end of June, Petopetal and Ernie's developer community had grown to around 14.7 million.
Speaker Change: afterservices to users
Robin Li: Meanwhile, hundreds of thousands of enterprises are now engaged with us, highlighting the expanding influence and utility of our ecosystem.
Speaker Change: we're doing the early stages of an exciting yet uncharted opportunity
Speaker Change: as we wemean
Speaker Change: the product we are listening to feedback from usersds creator and partners
Robin Li: Looking ahead, we will be patient. We will continue to iterate on Ernie's capabilities to, ensure we stay at the forefront of this new AI wave. By leveraging real world applications and feedback from our ecosystem partners, we gain valuable resources and insights to continuously refine our models.
Speaker Change: we're focused on our long-term system notability over four-term onvalation
Robin Li: The synergy among our vibrant ecosystem, advanced model capabilities, and application driven approach will continue to strengthen our leadership in this rapidly evolving AI landscape.
Speaker Change: we will be patient and make sure that we sp the right foundation for new ecosystem and closely track user matrix
Operator: Ladies and gentlemen, we will now begin the question and answer session. If you wish to ask a question, you will need to press star one on your telephone and wait for your name to be announced. If you wish to cancel your request, please press star two.
Speaker Change: we believe this approach well delivever with their value to our shareholders
Operator: Gate.
Alex Yao: Your first question comes from Alex Yao with JP Morgan. Please ask your question. Thank you, management, for taking a question.
Speaker Change: thank you
Speaker Change: your next clquestionation comes from lincoln callong with goldman sas please ask your question
Operator: Thank you.
Gary Yu: Your next question comes from Gary Yu with Morgan Stanley.
Speaker Change: thank you management and taking my question so my question that's not the cow business so what are the driers behind this accelerating growth of co revenue this year how much of this incremental revenue how this year's a i related and how that's this contribute to the long-term payi call new grow
Robin Lee: I have a view on AI development. So first of all, how do you guys foresee the competition in AI evolving over the next two to three years? Will Baidu's current first mover advantage sustain, or will the technology gap among key players narrow over time? How do you guys view the current competitive landscape compared to, for example, Kimi, Doubao, Baba and Tenzin? In addition to technology leadership, what other factors are crucial to build a successful ecosystem in the AI industry? Thank you.
Speaker Change: also what's the margin prove for the ai t compared to the traditional accart business
Speaker Change: so how should you think about thea intense pricing competition amounabout different players in market and ultimately with optimal longer-term for this rcture for our car business thank you
Gary Yu: Please ask your, questions.
Lincoln: higher lincoln thanks for your question ult so as you only may have a noticed our e revenue shifted from the yield be year declinine q three last year to a put growth in q four and then kept accelerating
Robin Lee: Hi, Alex, this is Robin. Competition will be fierce over the next two to three years. Other may think whoever can make money from it will be able to survive.
Gary Yu: Hi, thank you for the opportunity to ask questions.
Gary Yu: My questions are related to AI and search.
Robin Lee: Last year, when China's AI industry was still at the so-called war of foundation models period, I predicted that the industry would go through consolidation, and only a few foundation models would survive. We are seeing a strong demand for our Gen AI products, and I'm proud that Ernie has continued to strengthen his market lead. And I'm confident we will maintain our leadership going forward. Our decade-long investment in AI provided us with the first mover advantage and technological leadership.
Lincoln: so this quarter acount revenue grew by fourteen percent -yearover-year
Lincoln: the rreapcord book where was a riven by strong demand for our gi and across various industries
Lincoln: thanks to our reputation for having the most powerful and cost effective infrastructures in the mar pipe
Robin Lee: In the healthcare industry, Dr. Burmout is one of the greatest challenges.
Lincoln: sowe're deaning a rich condition across the well range of sectors from area addoctors like internet companies in online education providers
Robin Lee: By leveraging early speed and our model builder, we help a healthcare automation solution provider to train and fine tune an industry-specific model that enables automatically generated medical records for doctors. This solution significantly reduces the administrative burden on doctors, a major factor in burnout, and enhances clinical efficiency. After two months of model deployment, doctors using it were able to treat 50% more patients on average.
Robin Lee: In the recruitment industry, we have collaborated with the recruiting service company.
Lincoln: to the industries like automotive financial services and utilities
Speaker Change: leading commtees companies like a now state great sfx express and goman pension and insurance in also the rising
Robin Lee: You know, we were the first public company globally to launch a GPT-like model in March of last year. And since then, we keep leveraging our self-developed four-layer AI architecture for model upgrade. Last October, we launched the Ernie 4.0 that's China's first GPT-4 type model. And this June, we launched the Ernie 4.0 Turbo, which offers superior capabilities over 4.0. We will continue to draw out upgraded versions of our flagship foundation models. And our application-driven approach further expands our competitive advantage and set us apart from the others.
Speaker Change: air products such as social technology so the
Speaker Change: or intended to our form in -depth punishable withus leveraging our industry leading pa capabilities
Gary Yu: So what kind of progress has been made in renovating traditional search using Gen AI and the large language models?
Speaker Change: so it's robin just mentioned a key revenue driver for a cloud is the fast growing gi and foundation model religed revenue
Speaker Change: it has increased from about five percent of the total ai cloud revenue in that q four last year to our almost nine percent this quarter
Speaker Change: to me the devver niece of our customers we offer a well range of products and solutions
Speaker Change: so for those seeking modeal training and hosting services with provide of the most coough-effective and a powerful be infrastructure
Robin Lee: We believe that foundation models hold little practical value without real applications built up on them, widely used by users and customers. So I mentioned earlier that currently Ernie handles about 600 million API calls and generates about 1 trillion tokens every day, which we believe is the highest in China. We have been heavily using Ernie to renovate our own products, like Baidu Search or Baidu Vancouver, into AI native applications on a large scale.
Speaker Change: this quarter we have observed a newable increase in spending on our public cloud from gpu customers
Speaker Change: reflecting there are high satisfaction in prorossing years
Speaker Change: so so this strong customer endorsement wereinforces our confidence in continuing to gain our market share
Speaker Change #100: for the customers who are looking to a horness the powerof l end so ap course we oferred the earning family which includes some of the most powerful l and avable in the channel market
Robin Lee: And this approach has helped us gain valuable insights from users and continuously supported the upgrade of Ernie. And this experience allows us to create better solutions for our external customers when they use Ernie. At the same time, we have been working really hard to make Ernie increasingly accessible and affordable, so it can be used widely to address real world needs. We continue to lower the cost of model inference, expand our model portfolio, and develop two cases for model building and app building so that our customers, partners, developers can leverage the power of Ernie more efficiently and effectively.
Speaker Change #100: mass platform also provides comprehensive two kids for easy thank uning of costand mordeals and the development of a applications tailored to our specific needs
Speaker Change #101: we see the lm engger market is stilling is very early stes
Speaker Change #102: we're educating the market and making ai accessible of key
Speaker Change #103: on top of the model capabilities and eily use where adoption also depends on affordabilityam
Robin Lee: By leveraging early capabilities for API costs, our customer has significantly upgraded the matching process between job descriptions and resumes. The smart matching upgrade has reduced the labor cost in this process by over 50%.
Speaker Change #103: so are two encourage usage we have made eps cos free for three languweed earning models and cut prress
Speaker Change #104: for a fxsyim mols
Speaker Change #104: so this has led to a rapid increase in earnings p co
Speaker Change #105: the stor in justdecid the total number of dl a cost has searched from around two hundred million in minmade to over six hundred million which means over one threeill bas tokens
Robin Lee: And this application-driven approach has helped us create a dynamic and expanding ecosystem. By the end of June, Petal Petal and Ernie developer community had grown to around 14.7 million. Meanwhile, thousands of thousands of enterprises are now engaged with us, highlighting the expanding influence and utility of our ecosystem.
Speaker Change #105: we believe with the demand for earning su ap co we will continue to grow rapidly
Speaker Change #106: given earnings superior and continuously improving efficiency imforvements
Speaker Change #107: for traditional cloud services by cross-selling cpu cloud services to our existing g pu cloud of customers
Speaker Change #107: we have noticed a significant increase in the cpuel cloud revenue from these customers
Robin Lee: Looking ahead, we will be patient. We will continue to iterate on Ernie's capabilities to ensure we stay as a forefront of this new AI wave. By leveraging real world applications and feedback from our ecosystem partners, we gain valuable resources and insights to continuously refine our model. The synergy among our vibrant ecosystem, advanced model capabilities, and an application-driven approach will continue to strengthen our leadership in this rapidly evolving AI lens.
Robin Lee: Thank you.
Speaker Change #107: so theall we are pretty confident that our a cloud revenue growth will maintain strong momentum in the coming quarters and capitalize on a significant long-term growth optimunitiesves
Speaker Change #107: so on their profit side we are committed to a sustainable and housy growth
Speaker Change #107: so our ai cloud business is healthier than ever in the consistently grourip generating non-gaap operating profit and we have been that improving the margins
Speaker Change #108: so regarding the price competion issue your question we are know that gi in howhan market is stillue its various stage and accessibility with superior price andinforments is essential for growing the market
Gary Yu: Your next question comes from Gary Yu with Morgan Stanley. Please ask your questions. Hi, thank you for the opportunity to ask questions. My questions are related to AI and search. So what kind of progress has been made in renovating traditional search using GNI and the last-language models? What kind of KBI is the company monitoring for AI and search, and have you have served any changes in user behavior, since implementing AI in the search business?
Speaker Change #109: well we are working very hard to improve the capbabability of earning models we are also continuously lowering the inference cost
Gary Yu: I'm related to that. What is the role of applying for AI and search moving forward? And it puts specific KBI that the company aimed to achieve by the end of the year. For example, you know, certain percentage of search we saw generated by AI by the end of this year by in coming years. Thank you.
Speaker Change #110: conidering the sk effect as a leading player in the market i believe the way we'll kick growing our much share
Speaker Change #111: and also with the scale i believe with the normalized margin for our ga rel ary cloud business should exceed d of traditional cloud business
Speaker Change #112: thank you
Thomas Chong: next question comes from thomas chong with jay phrase please ask youra question
Thomas Chong: icould even ense management and ic king my question can management quantified the impact of the soft advertising revenue isdistinguishing between weakk mcall the canadgicalization impact on generative a implementation as well as the competition particularly in search
Speaker Change #114: separ what is the core advertising growth for the second half this year if we expect to see a turn alaround by year and is it more driven by better medical or better monetization in ai looking at two andtwenty five and beyond what the sustainable bothth rate thank you
Robin Lee: Yeah, we are very committed to using cutting-edge AI technologies to transform our search capabilities and blend search with speed in one app. So KPI is related to Baidu app, well, override search-only KPI. In the second quarter, we accelerated the pace of renovating our search by leveraging Ernie. Our primary goal is to improve user experience, offer more capabilities, and boost user engagement on Baidu. As highlighted in my opening remarks, we made several important progress.
Gary Yu: What kind of KPI is the company monitoring for AI in search?
Gary Yu: And have you observed any changes, in user behavior since implementing AI in the search business?
Speaker Change #115: yeah i found mentioned in my opening remarks our advertising businessis currently facing pressure costed by a accombination of external factors and our proactive efforts to accelerate eagdriven renovation of search
Gary Yu: Related to that, what is the rollout plan for AI in search moving forward?
Gary Yu: Any specific KPI that the company aim to achieve by the end of the year?
Speaker Change #116: externally macroweekness and compebititing pressure the primary challenges
Robin Lee: It is measured by total working hours while maintaining high quality results.
Speaker Change #117: on the macro front
Robin Lee: This automation allows the customer to expand its services to a broader client base, thereby increasing revenue generating capability, and demonstrating the tangible benefits of integrating earnings into its operations.
Robin Lee: First, we are providing smarter search results at an expanding pace. Currently, 18% of search results are provided by GNI. This results are more comprehensive and relevant than traditional search results. Offering information that was previously unavailable or scattered around different places. Furthermore, GNI enables infinite supply of constant on-demand and can effectively catch the decline of the open web system. Second, when GNI results can already meet user needs from the current query, we will recommend other contents or services just like a user would find in a typical new speed context.
Speaker Change #118: the recovery of consumer spending had been sluggish
Speaker Change #118: causing many advertisers especially the small in medium side ver heyer who spend haveily rely on offline activities to adopt a very cautious approach to their as spending
Speaker Change #118: certain verticals such as real estate and its related sectors franchising and automobile have been particularly weak
Speaker Change #119: inaddition to the chinese scener uusers behavior has changed it loted
Speaker Change #119: users by increasingly concentrating their time a selected a few leading short video and sohoto media platforms
Speaker Change #119: the user time spend share change can be a proxy into the loss off our online advertising been
Robin Lee: This will enhance user time spent and retention. Third, the new interactive function in Baidu search allows users to refine their questions and follow-off through multi-round conversations, reshaping how users express their needs and making the search experience more natural and efficient. Additionally, I would like to point out that we are rapidly rolling out Ernie agents with things search results to provide intelligent assistance to our users. By opening up our platform, developers can create a wide range of agents tailored to meet the unique needs of each and every user.
Gary Yu: For example, you know, certain percentage of search result generated by, AI either by the end of this year or by in coming years.
Speaker Change #120: and as i mentioned earlier er eighteen percent how our search suwithoutts contained jmy i content
Speaker Change #121: which is not manis at this time
Speaker Change #121: this number well inquase over time
Speaker Change #121: of course our aggressive rollout of geneny acuuring of the monthtiation system partly offset the revenue sacrificice here
Speaker Change #121: well this are all the big challenges for the time being we strongly believe technological innovation will be the key for our future success
Speaker Change #122: specifically we need to focus on what we can acve shape and influence
Robin Lee: Our vision is to create an ecosystem where many developers contribute to building a robust network of agents, collectively enriching the user experience and up-perform the most powerful frontier models, addressing a broad spectrum of challenges and opportunities. With all the efforts mentioned, the types of search queries on Baidu have become more diversified. Users are increasingly posing complex questions seeking comprehensive and in-depth answers from Baidu search. While this demands were once challenging to meet, they can now be efficiently addressed.
Speaker Change #123: elllaborating our strenges in the resources
Speaker Change #123: to se new opportunities with our reach
Speaker Change #123: in a fourter seiten ball growth over time
Speaker Change #123: so to this end we are renovating our entire mobile ecosystem with janary i and
Speaker Change #123: for the efward highting system the renovations are already showing some results we talk about this during the
Speaker Change #123: previous quarterters the incremental ad revenue generated data by ju i and l l continue to grow in q two this is mainly due to the improvement
Speaker Change #123: in the montiation system
Speaker Change #123: which have improved as conversion and efficiency
Speaker Change #123: and on top of that we actively exploring opportunities for earning agents to further enhance our advertising capabilities
Robin Lee: On top of that, users are also using Baidu to create content such as writing and editing text and images. Generally, I is transforming Baidu search into a multi-functional creative platform offering a wider range of services to users.
Speaker Change #123: however using ai to upgrade our existing advertising system is just a near-term debt we believe that the two long-term value light in the renovation of our consumer facing ps
Robin Lee: We are still in the early stages of an exciting yet uncharted opportunity. As we remain in the product, we are listening to feedback from users, creators, and partners. We are focused on long-term sustainability over short-term monetization. We will be patient and make sure that we set the right foundation for a new ecosystem and closely track user metrics. We believe this approach will deliver greater value to all our shareholders.
Speaker Change #124: with over twenty years of experienceing channel search in our industry-leading capabilities in generually i and am
Robin Lee: The customer is very satisfied with the result and is now exploring new collaboration opportunities.
Robin Lee: Thank you.
Speaker Change #124: by with uniquely positioned to innovate and transform genoc search into ai search
Robin Lee: In the public service sector, we utilize the early speed to deliver tailored, personalized, and scalable services, addressing challenges like limited resources and efficiency needs. This is particularly valuable in China's grassroots public services, where shortages in manpower make large-scale personalized services challenging. We have collaborated with the customer to help over 6,000 villages in China improve a public service, offering citizens more personalized and efficient support.
Robin Lee: Since the large-scale launch of the service in April, its daily usage has surged over 15 fold to more than 2 million times per day.
Speaker Change #124: this transformation is not just about enhancing the traditional surarch experience it's about redefining what search can do for users
Robin Lee: Helping locals handle tasks such as household registrations, social security inquiries, and tax declarations.
Speaker Change #125: things q two we have accelerated data the pace to renovate by the search shielding earning
Robin Lee: Internally, we have accelerated the renovation of title search with Ernie, significantly enhancing our consumer-facing products at large scale.
Speaker Change #125: new features such as ai generally arch without angeractive features missing search
Lincoln Kong: Your next question comes from Lincoln Kong with Goldman Sachs. Please ask your question. Thank you, manager, for taking my question. My question is about the cloud business. What are the drivers behind this accelerating growth of cloud revenue this year? How much of this incremental revenue will cloud this year's AI related and how does this contribute to the long-term AI cloud revenue growth? Also, what's the margin for the AI cloud compared to the traditional cloud business? How should we think about the intense pricing competition among different players in the market?
Speaker Change #125: and earning agents have been rolling out rapidly
Speaker Change #125: this renovation has high potential but is not something that well be completed in a short time
Speaker Change #125: during this process there are may be necttied impact on revenue but we are confident that the result will be a completely different new ai ecosystem
Speaker Change #125: one that is much more versatilel productive and user pendding
Speaker Change #125: so in may take some time before we see are noticeable revenue contribution from the i enhancements saying our search product
Robin Lee: Generative AI is transforming user search experience, pushing beyond traditional boundaries.
Rong Luo: [inaudible] So beginning a recognition across a wide range of sectors, from early adopters like internet companies and online education providers to the industries like automotive, financial services and utilities. Leading companies like Nasid Great, SFX Prize, and Guomin, Pension and Insurance, and also the rising AI startups, such as Shenzhou Technology, so they all intend to form in-depth partnership with us, leveraging our industry leading AI capabilities. So let's Robin just mention, a key revenue driver for AI cloud is the fast-growing Genie AI and Foundation model-related revenue.
Speaker Change #125: but we are confident that focusing on creating longlasting useth of value rather than short-term revenue well ultim equate tremendous value and drive sustainable revenue growth
Robin Li: Thank you.
Robin Li: Yeah, we are very committed to using cutting edge AI technologies to transform our search, capabilities and blend search with speed in one app.
Speaker Change #126: thank you
Miranda Z: your exclation comes from mirandaz z with b of america securities please ask youra question
Robin Li: So KPIs related to Baidu app will override search only KPIs.
Robin Li: In the second quarter, we accelerated the pace of renovating our search by leveraging early.
Robin Li: Our primary goal is to improve user experience, offer more capabilities, and boost user engagement on Baidu.
mirandaz z: thank you manageurment for taking my questions so my question is about the a i agent in a monetization okay company explay how the a i agcient is actually used in your research experience and monetization
Robin Li: As highlighted in my opening remarks, we've made several important progresses.
Robin Li: First, we are providing smarter search results at an expanding pace.
Robin Li: Currently, 18% of search results are provided by Gen AI. These results are more comprehensive and relevant than traditional search results, offering information that was previously unavailable or scattered around different places.
Robin Li: Second, when Gen AI results can already meet user needs from the current query, we will recommend other contents or services just like a user would find in a typical newsfeed context. This will enhance user time spent and retention.
Robin Li: Furthermore, Gen AI enables infinite supply of content on demand and can effectively, catch the decline of the open web system.
Robin Li: Third, the new interactive function in Baidu search allows users to refine their questions, and follow up through multi-round conversations, reshaping how users express their need and making the search experience more natural and efficient.
Speaker Change #129: how do you drive the adoption of these a agent especially amount as inany deficises which have less data and v check capability what kind of runup pace doue expect by the end of this year for the ai agents
Speaker Change #129: and also can management give us more color on the recent development in a e monetization especially the progress on the cpc to the cpss model the advertises takeick that on the effectiveness and the future budget allocation on a biital paform
Robin Lee: Since the second half of last year, we have been testing early-powered features on a small scale.
Robin Li: Additionally, I would like to point out that we are rapidly rolling out early agents within, search results to provide intelligent assistance to our users.
Robin Li: By opening up our platform, developers can create a wide range of agents tailored to meet the unique needs of each and every user. Our vision is to create an ecosystem where many developers contribute to building a robust, network of agents, collectively enriching the user experience and outperform the most powerful frontier models, addressing a broad spectrum of challenges and opportunities.
Robin Li: With all the efforts mentioned, the types of search queries on Baidu have become more diversified. Users are increasingly posing complex questions, seeking comprehensive and in-depth answers from, Baidu Search. While these demands were once challenging to meet, they can now be efficiently addressed.
Robin Li: On top of that, users are also using Baidu to create content, such as writing and editing text and images.
Speaker Change #130: ll thank you so much
Robin Li: Genium AI is transforming Baidu Search into a multifunctional creative platform, offering a wider range of services to users.
Speaker Change #131: yeah there are more than ten thousand earning agents but by now on our platform
Robin Lee: This quarter are focused as shifted to substantially amplifying search capabilities with advanced AI features.
Speaker Change #132: the beauty of or any agents is that anyone can quitate them whether you are user a quickator a service provider an advertiser or a company
Rong Luo: It has increased from about 5% of the total AI cloud revenue in a queue for last year to almost 9% this quarter. To make the diverse needs of our customers, we offer a wide range of products and solutions. So for those seeking model training and hosting services, we've provided the most cost-effective and powerful AI infrastructure. This quarter, we have observed a notable increase in spending on our public cloud from a GPU customers, reflecting their high satisfaction in trusting theirs.
Speaker Change #133: the barrier to entry very low it's really important to understand that the value of earning agents extends far beyond just scream a two or for manattiization
Speaker Change #133: they will significantly improve the user experiencing our mobust the ecosystem
Speaker Change #133: a recent example is a lawyer agent created and distributed on our platform
Speaker Change #134: a user had a conflict with a law enforcement officer and started this consultation with this agent
Speaker Change #135: we had item t thousand and eighty-eight rounds of conversation and lasted for five hours
Rong Luo: So this strong customer endorsement reinforces our confidence in continuing to gain market share. For the customers who are looking to harness the power of LNs, so API course, we offer the Ernie Finally, which includes some of the most powerful LNs available in the China market. Matt's platform also provides comprehensive two-kits for easy thanks-uning of cost and models and the development of applications tailored to us, especially these.
Speaker Change #136: as car to imagine a real human lawyer would have the time and patients to talk to potential client like that
Speaker Change #137: but the benefits for the user for the advertising law firm and for our platform as self evident in engl this case
Robin Li: We're still in the early stages of an exciting yet uncharted opportunity.
Speaker Change #138: today or ningthe agent is steel in its early stages and they evolving quickly but we're glad to see that advertiser by using earning agent
Speaker Change #138: um
Rong Luo: We see the LLM and Genie AI market is still in its very early stages. We're educating the market and making AI accessible or key. On top of the model capabilities and ease of use, wide adoption also depends on affordability. So to encourage usage, we have made API course free for three late-week early models and cut price for flagship models. So this has led to a rapid increase in the early API course.
Speaker Change #139: to get better salesie conbustions than then the traditional search has
Robin Li: As we reinvent the product, we are listening to feedback from users, creators, and partners.
Speaker Change #139: as i mentioned before more and more advertiserd especially in feuds like legal education and b to be starting to use earning engent
Speaker Change #139: this areas often require long decision making time and personalized presales consultation
Speaker Change #139: early agent with its interactive feature is getting better and better at addressing this kind of need
Rong Luo: Sorting just that, the total number of daily API course has searched from around 200 million in mid-May to over 600 million, which means over one trillion based tokens. We believe that the demand for Ernie's API course will continue to grow rapidly, given Ernie's superior and continuously improving efficiency and performance. For traditional cloud services, by cross-selling CPU cloud services to our existing GPU cloud customers. We have noticed a significant increase in the CPU cloud revenue from these customers.
Speaker Change #139: plus as ear agent becomees mararty it has the potential to drive direct cells and even more value for our abverththirs
Robin Lee: Currently, 18% of search without pages contain generated content. Up from 11% in mid-May.
Speaker Change #139: therefore moving us from a cpuc model to a more effective cs model
Robin Lee: This AI generated search without delivering more accurate and direct answers, enhancing content quality and providing previously unattainable information. This improvement has led to increased user satisfaction and engagement. As more users are now turning to Baidu Search for more complex queries, increasing Baidu Search versatility.
Robin Li: We're focused on long-term sustainability over short-term monetization.
Robin Li: We will be patient and make sure that we set the right foundation for a new ecosystem and closely track user metrics.
Speaker Change #139: we believe earning agent has potential to become the new weestite system in the ai erra
Operator: We believe this approach will deliver greater value to all our shareholders.
Speaker Change #139: early stage or any agents might not always improve the user experienceced right the way but as they start providing withing interactions
Speaker Change #139: boosting engagement
Speaker Change #139: in offering personinalizedines that services monpition opportunities well naturally followed
Rong Luo: So overall, we are pretty confident that our AI cloud revenue grows. We have maintained strong momentum in the common quarters and capitalized on a significant long-term growth opportunities. So on the profit side, we are committed to sustainable and healthy growth, so our AI cloud revenue is healthier than ever in the consistently-generating non-gap operating profit. And we have been improving the margins. So regarding the price-compatient issue in your question, we all know that Geniei and our own market is still in this very area stage.
Operator: Thank you.
Lincoln Kong: Your next question comes from Lincoln Kong with Goldman Sachs.
Speaker Change #139: thank you
Lincoln Kong: Please ask your question.
Speaker Change #139: next question comes from alicia yes west city group please ask your question
Lincoln Kong: Thank you, manager, for taking my question.
Lincoln Kong: My question is about the cloud business.
Lincoln Kong: What are the drivers behind this accelerating, growth of cloud revenue this year?
Speaker Change #140: hi thank you good evening management thanks for taking about questions have a questions on the margin
Alicia: so bythe core continue to achieve the expensions in the operating margins in the second quarter just wondering if this expansion sustainable what is your outlook on the profit margin trend for by the core in the highire year two thousand and twenty four thank you
Rong Luo: And accessibility with superior price and performance is essential for growing the market. While we are working very hard to improve the capability of early models, we are also continuously lowering the inference cost. Considering the skill effect, as leading player in the market, I believe we will keep growing our much share. And also with the skill, I believe that the normalized margin for Geniei, really to the cloud business, should exceed that of traditional cloud business. Thank you.
Lincoln Kong: How much of this incremental revenue in cloud this year is AI-related, and how does this contribute to the long-term AI cloud revenue growth?
Lincoln Kong: Also, what's the margin profile for the AI cloud compared to the traditional cloud business?
Speaker Change #142: of a question and limme ary question as you know that we don't give the guidance for quares hat by the try to give you some inight into our business priorities and what could they mean for cast e benes
Lincoln Kong: How should we think about the intense pricing competition among different players in the cloud?
Lincoln Kong: Ultimately, what is the optimal longer-term profit structure for our cloud business?
Speaker Change #143: for air clou we continue to fac ll the look margin businesses so we can continue generating operating profit and markets on a longoner basiness
Speaker Change #143: looking to a longter the normalized modeins follow ai related cloud businesses should be higher than a lacy cropces
Thomas Chong: Your next question comes from Thomas Chong with Jay Freys. Please ask your question. Hi. Good evening. Thanks, management, for taking my question. Can management quantify the impact of the soft advertising revenue? Distinguishing between week, month, or categorization impact from generative AI implementation, as well as the competition, particularly in search, separately? What is the core advertising growth for the second half this year? If we expect to see a turn around by year, and is it more driven by better macro or better monetization in AI, looking at 2025 and beyond what the sustainable growth rate. Thank you. Yeah.
Speaker Change #144: and this kind of revenue today's are that percent of upper to the oud revenue for intelligent driving and the other growth initiatives where p to reduce our lotes
Robin Lee: AI generated search without may reduce app impressions and therefore have a negative short-term impact on monetization, but they provide significant value to our users. By prioritizing long-term user experience over immediate revenue and profits, we see the strategic implementation of generated search as essential to driving future success.
Speaker Change #144: and but particularly are will in having druriving businesses with the operational efficient gs and try to reach a your improvements and global tax busence
Lincoln Kong: Thank you.
Speaker Change #145: overall we believe that high-quality growth and the investments should be weare balanced
Dou Shen: Hi, Lincoln.
Speaker Change #146: for many years now we have delivered a solid track record of the top line revenue growth and
Dou Shen: Thanks for your question.
Speaker Change #146: where a resolute cost discipling and intent to continue building
Dou Shen: This is Dao.
Speaker Change #147: our futureures only spaces somehigh of course theyll investment runs ahead of the high colleity groups
Dou Shen: As you may have noticed, our AI cloud revenue, shifted from a year-over-year decline in Q3 last year to a positive growth in Q4 and then kept accelerating. This quarter, AI cloud revenue grew by 14% year-over-year. The rapid growth is driven by strong demand for GNI and LLMs across various industries. Thanks to our reputation for having the most powerful, and cost-effective AI infrastructures in the Mars platform.
Dou Shen: So we're gaining recognition across a wide range of sectors, you know, from early adopters, like internet companies and online education providers, to the industries like automotive, financial services, and utilities.
Dou Shen: Leading companies like State Great, SF Express, and Guangming Pension and Insurance, and also, the rising AI startups, such as Shengshu Technology.
Dou Shen: So they all intend to form in-depth a partnership with us, leveraging our industry-leading AI, capabilities.
Speaker Change #147: we continue to operate our lacy such businesses new challenging environment that pressures our revenue
Rong Luo: [inaudible] The user time spend share change can be a proxy to the loss of our online advertising business. And as I mentioned earlier, 18% of our search results contain genuine eye content, which is not monetized at this time. This number will increase over time. Of course, our aggressive raw-out of genuine eye features in the monetization system, partly offset the revenue sacrifice here. Well, these are all the big challenges for the time being.
Speaker Change #148: to tangle with kind of of challenges is very important to focus on offering uses it impro improving experiences
Robin: in particular whole new er bed such experience for bythe search as robin has mentioned just now this quarter will have ceority the renovations are by those seual waste earning you short and we have seen significant increasing a i generative such results
Dou Shen: So as Robin just mentioned, a key revenue driver for AI cloud is the fast-growing gene, AI and foundation model-related revenue. It has increased from about 5% of the total AI cloud revenue in Q4 last year, to almost, 9% this quarter.
Dou Shen: To meet the diverse needs of our customers, we offer a wide range of products and solutions.
Dou Shen: So for those seeking model training and hosting services, we provide the most cost-effective, and powerful AI infrastructure.
Dou Shen: For the customers who are looking to harness the power of LMs through API calls, we offer, the early funneling, which includes some of the most powerful LLMs available in the China market.
Dou Shen: This quarter, we have observed a notable increase in spending on our public cloud from, GPU customers, reflecting their high satisfaction and trust in us.
Dou Shen: MADS platform also provides comprehensive toolkits for easy fine-tuning of custom models, and the development of applications tailored to specific needs.
Dou Shen: So this strong customer endorsement reinforces our confidence in continuing to gain market, share.
Dou Shen: We see the LLM and gene AI market is still in its very early stages.
Dou Shen: We're educating the market and making AI accessible or key. On top of the model capabilities and ease of use, wide adoption also depends on affordability. So to encourage usage, we have made API calls free for three lightweight early models and, cut price for flagship models. So this has led to a rapid increase in earning API calls. As Robin just said, the total number of daily API calls has surged from around 200 million, in mid-May to over 600 million, which means over 1 trillion base tokens.
Speaker Change #149: eleven percentcess ple of all such result pages contain the generative contents you may itmay this percentagegy has to rise cent eighteen percent one a up enough
Dou Shen: We believe that demand for earnings through API calls will continue to grow rapidly, given, earnings superior and continuously improving efficiency and performance.
Dou Shen: For our traditional cloud services, by cross-selling CPU cloud services to our existing GPU cloud, customers. We have noticed a significant increase in CPU cloud revenue from these customers.
Dou Shen: So overall, we are pretty confident that our AI cloud revenue growth will maintain strong momentum in the coming quarters, and capitalize on significant long-term growth opportunities.
Dou Shen: So on the profit side, we are committed to sustainable and healthy growth.
Dou Shen: So our AI cloud business is healthier than ever, consistently generating non-gap operating profit, and we have been improving the margins.
Dou Shen: Considering the skill effect as a leading player in the market, I believe we will keep growing our market share.
Dou Shen: So regarding the price completion issue in your question, we already know that GeneAI and LLM's market is still in its very early stage, and accessibility with superior price and performance is essential for growing the market. While we are working very hard to improve the capability of early models, we are also continuously lowering the inference cost.
Dou Shen: And also with the skill, I believe the normalized margin for GeneAI related to cloud business, should exceed that of a traditional cloud business.
Speaker Change #150: the scaling up our generating such results and our new features to require a lot more competing capacity
Speaker Change #151: and you usalso interram more in margin al conversations
Speaker Change #151: there are more complicated patr
Speaker Change #151: and they use the new contemrate creation capabilities within sucharch
Speaker Change #151: all of these raised the total cast as the more computing parties needitty while the new a search features has now a peat money has now
Speaker Change #152: so early a celerating a and made native transformation of such will youunnotably put some pressures on our margins in the shortten
Speaker Change #153: so we'll reduce our part of an efficient ai frasstructures and our technique but is to make efficientcy against over time thank of so much for cus
Operator: Thank you.
Operator: Your next question comes from Thomas Chong with Jefferies.
Rong Luo: We strongly believe technological innovation will be the key for our future success. Specifically, we need to focus on what we can actively shape and influence, leveraging our strengths and resources to see new opportunities within our reach and a faster system for growth over time. So to this end, we are renovating our entire mobile ecosystem with genuine eye and earned. For the advertising system, the renovations are already showing some results. We talked about this during the previous quarters.
Thomas Chong: Please ask your question.
ways young: next question comes from ways young with ubbs
Speaker Change #155: please ask your question
Thomas Chong: Hi, good evening.
Thomas Chong: Thanks, management, for taking my question.
Thomas Chong: Can management quantify the impact of the soft advertising revenue, distinguishing between weak macros, the categorization impact from generative AI implementation, as well as the competition, particularly in search, separately?
ways young: sure thank you management goodevening thank you for takinging my question my question is regarding the robo taxi business so with government support on accelerating the robo taxi license approval more cities how would that translateated by do rbo taxi opportunities and also cast associated to that
Speaker Change #156: and what are the targets for a pollogo this year and for the next few years how do wehave plan to reapplicate will hand success nationwide a example do you consider partnership or collaboration for what taxic future development and lastly could manage share your view on the comppetitive landscape change in the roprotexti market thank you
Rong Luo: The incremental ad revenue generated by genuine eye and LLM continue to grow in Q2. This is mainly due to the improvement in the monetization system, which have improved ad conversion and efficiency. And on top of that, we actively exploring opportunities for early agents to further enhance our advertising capabilities. However, using AI to upgrade our existing advertising system is just a near-term step. We believe that the true long-term value lies in the renovation of our consumer-facing products.
Thomas Chong: What is the core advertising goal for the second half this year?
Speaker Change #157: yeah being investing in on humps driving technology for over a decade guided by steady focus and our labghborouring patients
Speaker Change #158: robert taxi servyses have long been considered as the brakthroughule so advanced that no one can predict exactly how long it well take to bring it to four frion
Thomas Chong: If we expect to see a turnaround by year end, is it more driven by better macro or better monetization in AI?
Speaker Change #159: for a long time many doubted whether we or anyone else in this world would make it but we kept moving forward now as we see regional u you break even approaching our bition is' becoming a scalable reality
Thomas Chong: Looking at 2025 and beyond, what's the sustainable growth rate?
Robin Lee: Our AI investments are fostering deeper user interactions. New interactive features enable users to refine their questions through multi-round conversations and enhancing the overall user experience.
Rong Luo: With over 20 years of experience in general search and our industry-leading capabilities in genuine eye and LLM, by doing uniquely positioned to innovate and transform general search into AI search. This transformation is not just about enhancing the traditional search experience. It's about redefining what search can do for users. Since Q2, we have accelerated the pace to renovate by the search using learning. New features such as AI generated search without interactive features within search and early agents have been rolling out rapidly.
Robin Li: Thank you.
Robin Li: Yeah, as I mentioned in my opening remarks, our advertising business is currently facing pressure caused by a combination of external factors and our proactive efforts to accelerate the AI driven renovation of search.
Robin Li: Externally, macro weakness and competitive pressure are the primary challenges. So on the macro front, the recovery of consumer spending has been sluggish, causing many advertisers, especially the small and medium sized advertisers, who heavily rely on offline activities to adopt a very cautious approach to their ad spending.
Robin Li: Certain verticals, such as real estate and related sectors, franchising and automobile, have been particularly weak.
Robin Li: In addition to this, Chinese internet users' behavior has changed a lot. Users are increasingly concentrating their time on a selected few leading short video and social media platforms.
Robin Li: The user time span share change can be a proxy to the loss of our online advertising business.
Speaker Change #159: our persistence even in the phase ofckdoubt is's not only driving progress it's shaping the future
Robin Li: And as I mentioned earlier, 18% of our search results contain GeneAI content, which is not, monetized at this time. This number will increase over time.
Robin Li: Of course, our aggressive rollout of GeneAI features in the monetization system partly, offset the revenue sacrifice here.
Robin Li: While these are all the big challenges for the time being, we strongly believe technological, innovation will be the key for our future success. Specifically, we need to focus on what we can actively shape and influence, leveraging, our strengths and resources to seize new opportunities within our reach and foster systems for growth over time.
Robin Li: We talked about this during the previous quarters.
Robin Li: So to this end, we are renovating our entire mobile ecosystem with GeneAI and EARN.
Robin Li: The incremental ad revenue generated by GeneAI and LLM continue to grow in Q2. This is mainly due to the improvements in the monetization system, which have improved, ad conversion and efficiency.
Robin Li: For the advertising system, the renovations are already showing some results.
Robin Li: And on top of that, we are actively exploring opportunities for early agents to further, enhance our advertising capabilities.
Robin Li: However, using AI to upgrade our existing advertising system is just a near-term step.
Operator: Thank you.
Robin Li: We believe that the true long-term value lies in the renovation of our consumer-facing products.
Operator: Your next question comes from Miranda Zhang with Bank of America Securities.
Robin Li: With over 20 years of experience in general search and our industry-leading capabilities, in GeneAI and LLMs, Baidu is uniquely positioned to innovate and transform general search into, AI search.
Robin Li: This transformation is not just about enhancing the traditional search experience, it's about, redefining what search can do for users.
Speaker Change #159: our dedication has made a pologo the world's largest to upha rright hillning service provider this is measured by number of dides
Robin Li: In Q2, we have accelerated the pace to renovate Baidu search using EARN. New features such as AI-generated search results, interactive features within search, and early, agents have been rolling out rapidly. This renovation has high potential, but it's not something that will be completed in a, short time.
Robin Li: During this process, there may be a negative impact on revenue, but we are confident that, the result will be a completely different new AI ecosystem. One that is much more versatile, productive, and user-friendly.
Robin Lee: Additionally, there is a significant increase in users utilizing text and image creation tools within search.
Robin Li: So it may take some time before we see a noticeable revenue contribution from the AI enhancements, in our search product.
Robin Li: But we are confident that focusing on creating long-lasting user value rather than short-term, revenue will ultimately create tremendous value and drive sustainable revenue growth.
Miranda Zhang: Please ask your question.
Miranda Zhang: Thank you, management, for taking my questions.
Speaker Change #159: our track wecord of
Speaker Change #159: safe operations have earned public and regulatory trust
Miranda Zhang: So my question is about the AI agent and the monetization.
Speaker Change #159: allowing us to expen servic this and reach more people
Miranda Zhang: So can company explain how the AI agent is actually used in your user experience and, monetization?
Speaker Change #160: we happy to see that more and more passengers are choosing a ho go not just for testing rs but as we are ago to option for daily communutity
Miranda Zhang: How do you drive the adoption of these AI agents, especially among the SME advertisers, which have less data and less check capability?
Miranda Zhang: What kind of run-up pace do you expect by the end of this year for the AI agents?
Miranda Zhang: And also, can management give us more color on the recent development in the AI monetization, especially the progress from the CPC to the CPS model?
Miranda Zhang: What are the advertisers' feedback on the effectiveness and the future budget allocation, on the Baidu platform?
Robin Li: Thank you so much.
Rong Luo: This renovation has high potential, but it's not something that will be completed in a short time. During this process, there may be a negative impact on revenue, but we are confident that the result will be a completely different new AI ecosystem, one that is much more versatile, productive, and user-friendly. So it may take some time before we see a noticeable revenue contribution from the AI enhancement in our search product, but we are confident that focusing on creating long-lasting user value rather than short-term revenue will ultimately create tremendous value and drive sustainable revenue growth. Thank you.
Robin Li: Yeah.
Speaker Change #160: we have stand as our largest operational area it's a plea of around four hundred hicles
Robin Li: There are more than 10,000 earning agents right now on our platform.
Robin Li: The beauty of earning agents is that anyone can create them, whether you are a user, a, creator, a service provider, an advertiser, or a company. The barrier to entry is very low.
Robin Li: It's really important to understand that the value of earning agents extends far beyond, just being a tool for monetization.
Speaker Change #160: offering twenty-four by sevenate service to nine million right stands there
Robin Li: It will significantly improve the user experience in our mobile ecosystem.
Robin Li: A recent example is a lawyer agent created and distributed on our platform.
Robin Li: A user had a conflict with a law enforcement officer and started this consultation with, this agent.
Robin Li: They had, I don't know, 288 rounds of conversation and lasted for five hours.
Robin Li: It's hard to imagine a real human lawyer would have the time and patience to talk to a potential, client like that.
Robin Li: But the benefits for the user, for the advertising law firm, and for our platform are self-evident, in this case.
Robin Li: Today, earning agents is still in its early stages and it's evolving quickly.
Speaker Change #161: our focus on cost reduction s innovation has also made our services increasingly affordable
Operator: Your next question comes from Alicia Yap with Citigroup.
Robin Li: But we are glad to see that advertisers are using earning agents to get better sales fee, conversions than the traditional search engines.
Alicia Yap: Please ask your question.
Robin Li: As I mentioned before, more and more advertisers, especially in fields like legal, education, and B2B, are starting to use early agent. These areas often require long decision-making time and personalized pre-sales consultation. Early agent, with its interactive features, is getting better and better at addressing, this kind of need.
Alicia Yap: Hi.
Robin Li: Plus, as early agent becomes smarter, it has the potential to drive direct sales and even, more value for our advertisers, therefore moving us from a CPC model to a more effective, CPS model.
Alicia Yap: Thank you.
Robin Li: We believe early agent has the potential to become the new website system in the AI era. Early stage early agents might not always improve the user experience right away, but, as they start providing greater interactions, boosting engagement, and offering personalized services, monetization opportunities will naturally follow.
Alicia Yap: Good evening, management.
Operator: Thank you.
Alicia Yap: Thanks for taking my questions.
Alicia Yap: I have a question on the margin.
Alicia Yap: So, Baidu Core continues to achieve the expansions in the operating margins in the second quarter.
Speaker Change #162: that is launching in an in q three of two thousand and twenty-two
Alicia Yap: Just wondering, is this expansion sustainable?
Robin Lee: This development opened doors for long-term value creation.
Alicia Yap: What is your outlook on the profit margin trend for Baidu Core in the entire year 2024?
Rong Luo: Thank you.
Rong Luo: Thank you for your question.
Rong Luo: Let me take your question.
Rong Luo: As you know, we don't give the guidance for quarters ahead, but I will try to give you, some insights into our business priorities and what could they mean for our costs and expenses.
Speaker Change #162: we've steadily increased the number of fully dabinic vehicles and rise and i mentioned that when per fully dargetics were right hing operation has been achieved in prulate the entire city
Operator: Your next question comes from Wei Xiong with UBS.
Rong Luo: For AI Cloud, we continue to phase out the low margin businesses so we can continue generating, operating profit and margins on a longer basis.
Operator: Please ask your question.
Rong Luo: Looking to the long term, the normalized margins for the general AI-related cloud businesses, should be higher than the legacy cloud businesses.
Wei Xiong: Sure.
Rong Luo: And this kind of revenue today is already 9% of our total cloud revenue.
Wei Xiong: Thank you, management.
Rong Luo: And for intelligent driving and the other growth initiatives, we aim to reduce our losses, and particularly our intelligent driving businesses with the operational efficiency gains and try to reach improvements in robotics business.
Wei Xiong: Good evening.
Rong Luo: Overall, we believe that high-quality growth and the investments should be well-balanced. For many years now, we have delivered a solid track record of the top-line revenue growth, and with a resolute cost discipline, we intend to continue building our futures on this basis. Sometimes, of course, though, the investment runs ahead of the high-quality growth.
Wei Xiong: Thank you for taking my question.
Rong Luo: We continue to operate our legacy search businesses in a challenging environment that pressures, our ad revenue.
Wei Xiong: My question is regarding the Robotaxi business.
Rong Luo: To tackle these kinds of challenges, it's very important to focus on offering the users, improved experiences.
Wei Xiong: So with government support on accelerating the Robotaxi license approval in more cities, how would that translate to Baidu's Robotaxi opportunities and also costs associated to that?
Rong Luo: In particular, a whole new AI-based search experience for Baidu Search, research.
Wei Xiong: And what are the targets for Apollo Go this year and for the next few years?
Rong Luo: As Robin has mentioned just now, this quarter we have a salary T, the renovations of Baidu, search with earnings.
Wei Xiong: How do we plan to replicate Wuhan's success nationwide?
Rong Luo: In the short term, we have seen a significant increase in AI-generated search results. 11% of all search result pages contain the generative contents in mid-May, and this percentage, has risen to 18%, 1A, as of now.
Rong Luo: The scaling of the AI-generated search results and other new features require a lot more, computing capacity, and users interact more in multiple conversations.
Rong Luo: They ask more complicated questions, and they use the new content creation capabilities, within search.
Rong Luo: All of these raise the total cost of the more computing power is needed, but the new AI, search features have not yet been monetized now. So the accelerated AI-native transformation of search will inevitably put some pressures, on our margins in the short term.
Rong Luo: So we will use our powerful and efficient AI infrastructures and our technical expertise, to make efficient gains over time.
Rong Luo: Thank you so much for your question.
Speaker Change #162: this have a significant impact on costs
Speaker Change #162: with our rc six vehicles being tested as scale we expectand continued cost reduction once rhissix officially put into operation
Alicia Yap: Your next question comes from Miranda Zhang with Bank of America Securities. Please ask your question. Thank you, management, for taking my questions. So my question is about the AI agent in the monetization. So can company explain how the AI agent is actually used in your user experience and monetization? How do you drive the adoption of this AI agent, especially among the SME advertisers, which have less data and less check capability? What kind of run-top pace do you expect by the end of this year for the AI agents?
Robin Lee: Last quarter, I mentioned that we planned to accelerate the distribution of earning agents. So far, we have seen more and more developers, advertisers and partners, together on our platform to develop innovative AI agents. The earning agent distribution within Baidu increased dramatically in July, exceeding 8 million daily, more than double the number in May.
Speaker Change #162: so please note our share up by haating market in muhan is still very small and iraround web percent and the sskcaling well beia gradual process and it could take many years
Robin Lee: Currently, the most frequently used agents by users for content creation, personal insights such as personality testing and tools such as translation and schedule planning.
Wei Xiong: For example, do you consider partnership or collaboration for Robotaxi future development?
Speaker Change #162: the potential of china's lightking marking is enormmat
Speaker Change #162: as ra the tax are inherently offers more affordable servourists we well naturally drive demand and openened up great new opportunities
Robin Lee: At the same time, we also see a joint of earning agents being more widely developed and applied both internally and externally. Our first college application assistant agent is one example of how earning agents are being distributed within Baidu to assist users with complex problem solving and decision making.
Alicia Yap: And also, can management give us more color on the recent development in the AI monetization, especially the progress from the CPC to the CPS model? What are the advertisers feedback on the effectiveness and the future budget allocation on the more than 10,000 early agents by now on our platform? The beauty of early agents is that anyone can create them, whether you are a user, a creator, a service provider, an advertiser, or a company.
Speaker Change #162: this is where innovation meets the v scale of the market
Speaker Change #162: and creating a powerful strenergy that will reshape the landscaave off transportation and beyond
Robin Lee: At the year after China's National College entrance examination.
Speaker Change #162: to an expanding population
Speaker Change #162: do
Speaker Change #163: also need to note that entering the robert taxing market requires deep believe in technology long-term investment
Robin Lee: University after checking exam results is crucial for the 10-million-plus candidates.
Speaker Change #163: and
Speaker Change #163: strong capital reserves
Wei Xiong: And lastly, could management share your view on the competitive landscape change in the, Robotaxi market?
Robin Li: Thank you.
Alicia Yap: The barrier to entry is very low. It's really important to understand that the value of early agents extends far beyond just being a tool for monetization. It will significantly improve the user experience in our mobile ecosystem. A recent example is a lawyer agent created and distributed on our platform. A user has had a conflict with a law enforcement officer and started this confrontation with this agent. We had 288 runs of conversation and lasted for five hours.
Speaker Change #164: a who goes stands out in this critical areas setting the benchmark or the industry
Robin Li: Yeah, we've been investing in autonomous driving technology for over a decade, guided by steady, focus and unlaboring patients.
Robin Li: Robotaxi services have long been considered as a breakthrough, so advanced that no one, can predict exactly how long it will take to bring it to full fruition.
Robin Li: For a long time, many doubted whether we or anyone else in this world would make it, but, we kept moving forward.
Robin Li: Now, as we see regional UU break-even approaching, our vision is becoming a scalable reality. Our persistence, even in the face of doubt, is not only driving progress, it's shaping, the future.
Robin Li: Our dedication has made Apollo Go the world's largest autonomous ride-hailing service provider. This is measured by number of rides.
Robin Li: The potential of China's ride-hailing market is enormous.
Robin Li: Our track record of safe operations has earned public and regulatory trust, allowing us to expand services and reach more people.
Robin Li: As Robotech inherently offers more, affordable services, it will naturally drive demand and open up great new opportunities.
Robin Li: We're happy to see that more and more passengers are choosing Apollo Go, not just for testing rides, but as their go-to option for daily commuting.
Robin Li: This is where innovation meets the vast scale of the market, and creating a powerful synergy that will reshape the landscape of transportation and beyond.
Robin Li: Wuhan stands as our largest operational area. It's a fleet of around 400 vehicles, offering 24 by 7 service to 9 million residents there.
Speaker Change #165: as oral question regarding to future band model currently all of our vehicles are own by us and we bear the hardbware capex and depreciating costs
Robin Li: As ApolloGo scales up, it will offer increasingly affordable, convenient, and much safer options to an expanding population.
Robin Li: Our focus on cost reduction through innovation has also made our services increasingly affordable.
Robin Li: Also need to note that entering the Robotech market requires deep belief in technology, long-term investment, and strong capital reserves.
Robin Li: Since launching in Wuhan in Q3 of 2022, we've steadily increased the number of fully driverless vehicles and, rides, and I mentioned that 100% fully driverless ride-hailing operation has been achieved in practically the entire city. This has a significant impact on costs.
Robin Li: ApolloGo stands out in this critical area, setting the benchmark for the industry.
Robin Li: As for your question regarding to future BIMS models, currently all of our vehicles are owned by us, and we bear the hardware, capex, and depreciation costs.
Robin Li: With our RT6 vehicles being tested at scale, we expect continued cost reduction once RT6 are officially put into operation.
Robin Li: So please note, our share of ride-hailing market in Wuhan is still very small at around 1%, and the scaling will be a gradual process and could take many years.
Robin Li: But we are open to various BIMS models and partnerships.
Speaker Change #165: but we are open to where its spin models and partnerships
Speaker Change #166: well our immediate to focus is still achieving uniber even in kcityity icative
Robin Li: While our immediate focus is still achieving UE breakeven in K-Cities. In the long run, we aim to innovate with asset-light models, like technology licensing and fleet operation, to capture the huge market opportunities, and ultimately benefiting society with better transportation solutions.
Robin Li: Thank you.
Speaker Change #166: in long long we aim to innoate its assetlight models like technology licensing and fleet operation
Rong Luo: It's hard to imagine a real human lawyer would have the time and patience to talk to a potential client like that, but the benefits for the user, for the advertising law firm, and for our platform, are self-evident in this case. Today, early agent is still in its early stages and it's evolving quickly, but we're glad to see that advertiser by using early agent to get better selfie conversions than the traditional search.
Speaker Change #166: to capture the huge market opportunities and ultimately benitting society with better transportation solutions
Operator: Ladies and gentlemen, that does conclude our conference for today.
Speaker Change #167: thank you
Speaker Change #168: ladies and gentlemen that does conclude our conference for today thank you for participating you may all disconnect
Robin Lee: In June, Baidu launched an innovative college application assistant agent, which is designed to meet personalized needs for selecting universities and majors. Following the National College entrance exam, the peak daily active users of the agent approached 2 million highlighting its significant impact and utility.
Rong Luo: As I mentioned before, more and more advertisers, especially in fields like legal, education, and B2B, are starting to use Ernie Agent. This area is often required long decision-making times and personalized pre-sales confrontation. Ernie Agent, with its interactive features, is getting better and better at addressing this kind of needs. Plus, as Ernie Agent becomes smarter, it has the potential to drive direct sales and even more value for our advertisers. Therefore, moving us from a CPC model to a more effective CPS model.
Robin Lee: While all these new features for more sophisticated user needs have not yet been monetized, they are transformative for us, for search, and a crucial for our future success. This strategic focus positions us to capture substantial long-term growth opportunities, transforming the usage of search and solidifying our leadership in the AI-driven search landscape.
Robin Lee: Our one-stop shaft for document creation is another example of how Baidu scales AI to address real-world problems and generate substantial value.
Robin Lee: As our most pioneering internal product to embrace generating AI and LLM, Baidu Wenku is now re-amping the benefits of its product renovation efforts.
Rong Luo: Ernie Agent, we believe Ernie Agent has the potential to become the new WebFact system in the AI era. Early stage Ernie Agent might not always improve the user experience right away, but as they start providing greater interactions, boosting engagement, and offering personalized services, monetization opportunities, will naturally follow.
Alicia Yap: Thank you.
Alicia Yap: Your next question comes from Alicia Yap, West City Group. Please ask your question. Hi, thank you.
Rong Luo: Good evening, management. Thanks for taking my questions. Have a question on the margin. So Baidu Core continue to achieve the expansions in the operating margins in the second quarter. Just wondering, is this expansion sustainable? What is your outlook on the profit margin trend for Baidu Core in the entire year 2024? Thank you.
Robin Lee: In the second quarter, Wenku's subscription revenue marked a year over increase in high teams.
Rong Luo: Thank you for a question. Let me hear a question. As you know, we don't give the guidance for quarter's hat. Baidu will try to give you some insights into our business priorities. What could they mean for our cost and expenses? For AI Cloud, we continue to face all the low margin businesses, so we can continue generating operating profit margins on a longer basis. Looking into the long-term, the normalized margins for the general AI-related cloud businesses should be higher than the Lexi cloud businesses.
Rong Luo: And this kind of revenue today is already not percent of our total cloud revenue. For intelligent driving and the other growth initiatives, we tend to reduce our losses, and particularly our intelligent driving businesses with the operational efficient gains, and try to reach your improvements in global tax savings. Overall, we believe that high-quality growth and the investments should be well balanced. For many years now, we have delivered a solid record of the top-line revenue growth and with a resolute cost-discipline and intend to continue building our futures on the basis.
Rong Luo: Sometimes, of course, there are investment runs at high-quality growth. We continue to operate our legacy such businesses in a challenging environment that pressures our revenue to tackle this kind of challenges. It's very important to focus on offering the uses, improving experiences, in particular a whole new AI base such experience for Baidu.
Operator: Thank you for participating.
Operator: You may all disconnect.
Operator: [music]
Robin Lee: AI is re-groupinating the platform with comprehensive content understanding and generation capabilities, while continuously improving in versatility.
Robin Lee: Since May, Baidu Wenku has launched new features resonating particularly well with young users. Such as crafting lengthy documents with tens of thousands of characters, building and editing PowerPoint presentations, and creating children's picture books with natural language guidance. The heightened engagement underscores Wenku's appeal to the next generation and signals promising future growth.
Rong Luo: Zhuang Li, Xiaomeng Zhuang Li, Xiaomeng Zhuang[inaudible] Xiaomeng Zhuang Li We're happy to see that more and more passengers are choosing a power goal, not just for testing rights, but as their go-to option for daily community. We'll handstand as our largest operational area with a fleet of around 400 vehicles offering 24 by seven service to 9 million residents there. Our focused on cost reduction through innovation has also made our services increasingly affordable.
Robin Lee: Moving from the digital realm to addressing changes and scaling AI in physical world, our decade-long dedication to autonomous driving innovation and long-standing investment is bearing fruit. Apollo Gaul, our autonomous ride hailing service, has recently achieved two significant breakthroughs.
Robin Lee: If establishing a robust foundation for making commuting more affordable and benefiting more people through technological innovations.
Robin Lee: First, things June 19.
Rong Luo: Since launching in Wuhan in Q3 of 2022, we've steadily increased the number of fully charging that 100% fully charged right-wing operation has been achieved in practically the entire city. This has a significant impact on costs. With our RT6 vehicles being tested at scale, we expect continued cost reduction once RT6 officially put into operation. So please note, our share of right-winging market in Wuhan is still very small and around 1% and the scaling will be a gradual process and it could take many years.
Robin Lee: Building our proven track record of safe operation Apollo Gaul has successfully transitioned to offering 100% fully driverless ride hailing services in June, practically the entire Wuhan Munisi Palace.
Rong Luo: The potential of China's right-winging market is enormous. Let's draw the tax that inherently offers more affordable services, it will naturally drive demand and open up great new opportunities. This is where innovation meets the vast scale of the market and creating a powerful synergy that will reshape the landscape of transportation and beyond. As Apollo goes scales up, it will offer an increasingly affordable, convenient and much affordable services to an expanding population. Also need to note that entering the robot taxi market requires deep believes in technology, long-term investment and strong capital reserves.
Robin Lee: This means all vehicles are now operating without the need for a human safety officer on board, a major death forward in making autonomous ride hailing bins commercially viable.
Robin Lee: Another milestone was the large scale open road testing of our six-generation autonomous vehicle, the RT-6.
Robin Lee: It could be a bad-waste watching solution, RT-6 is competitively priced at below $30,000 US dollars for mass production.
Robin Lee: After several testing, we plan to officially roll out RT-6 into our fleet, establishing a strong foundation for further substantial cost reduction in our political operations.
Robin Lee: All in all, our efforts to scale AI to solve real-world problems fully align with our broader vision of creating a more efficient, equitable, and sustainable future.
Rong Luo: Apollo goes stands out in this critical area, setting the benchmark for the industry. As for our question regarding future Benz model, currently all of our vehicles are owned by us and we bear the hardware, CapEx, and depreciation cost. But we are open to where with Benz models and partnerships. While our immediate focus is still achieving unique work even in case cities. In the long run, we aim to innovate as that light models like technology licensing and fleet operation to capture the huge market opportunities. And ultimately, benefiting society with better transportation solutions.
Operator: Thank you.
Robin Lee: Looking at how our efforts are yielding promising results, we believe we are well positioned to capitalize on the growing demand for AI-driven applications. By leveraging AI to address complex challenges, we can continue our journey of innovation and growth and create long-lasting value for our stakeholders.
Robin Lee: I want to thank our employees for their efforts and contributions and our customers and shareholders for their trust in us.
Robin Lee: Now, let me re-ask the key highlights for each business for the second quarter.
Operator: Ladies and gentlemen, that does conclude our conference for today. Thank you for participating. You may all disconnect.
Robin Lee: AI cloud revenue reached RB 5.1 billion, marking a consecutive acceleration to 14% over your growth while sustaining non-gab operating profitability in the second quarter. The strong growth is mostly attributable to the following two factors. First, Gen-AI-related revenue continued its robust momentum, accounting for nearly 9% of our total AI cloud revenue in Q2, up from 6.9% in the previous quarter.
Robin Lee: As more enterprises ink with Gen-AI and Foundation models into their daily operations, they increasingly come to us thanks to our reputation as China's most advanced and cost effective AI infrastructure providers and our excellent math platform for model training and inference.
Robin Lee: During the quarter, we further advanced our AI infrastructure management, intencing our ability to combine GPUs from more vendors for optimal training and hosting of models. This ensures flexibility, reliability and efficiency, positioning us to capture a larger share of the GPU cloud and LLM market.
Robin Lee: We also continued to develop the toolkit on our math platform for our customers and partners.
Robin Lee: Innebling the early family of models to deliver a superior price performance ratio.
Robin Lee: We're proud that our model builder now supports the full model development lifecycle, including data management, fine tuning, evaluation, optimization, and prompt engineering. A major upgrade has been the introduction of diverse high-grade training data sites, spanning from general to specialized industry tailored data sets. This enables efficient fine tuning and ensures high performance for industry specific applications while retaining strong general and capabilities.
Robin Lee: Also, thanks to our continuous refinement to app builder, the number of AI native apps on our cloud now runs into the hundreds of thousands. These applications span a wide array of industries and scenarios, from online education and e-commerce to public service sectors, and are integrated into both online platforms and smart devices.
Robin Lee: The second major driver of AI cloud revenue acceleration is cross-selling of our CPU cloud services to our GPU cloud customers.
Robin Lee: In Q2, we continue to observe an increase in CPU spending among our GPU cloud customers.
Robin Lee: We see that our strong brand recognition in GPU cloud is helping us win businesses in the CPU cloud industry.
Robin Lee: With brand new acceleration, AI cloud business continue to deliver positive non-gap operating profit and improved margins.
Robin Lee: Legacy cloud saw margin expansion and gen AI-related businesses are expected to have higher normalized margin compared to the traditional cloud service.
Robin Lee: Overall, we remain confident in the strong growth outlook for our AI cloud revenue and we aim to continue generating non-gap operating profit going forward.
Robin Lee: For mobile ecosystem, title course online marketing revenue declined by 2% year-over-year in the second quarter due to broader macroeconomic challenges, competition, and our aggressive AI driven search renovation.
Robin Lee: Key offline verticals such as real estate, franchising, and automobile remained subdued.
Robin Lee: Despite these challenges, we remain committed to optimizing our operations to maintain both a healthy margin and cash flow while accelerating the AI-nating transformation of our products.
Robin Lee: In Q2, incremental ad revenue from gen AI and LLM enhancements to our advertising system continue to grow quarter-over-quarter driven by the ongoing reconstruction of our monetization system and marketing cloud.
Robin Lee: We, Ignatian, Ernie Agent, as a compelling opportunity for our admins.
Robin Lee: Ernie Agent, as a potential to revolutionize our traditional CPC model into a more efficient CPS model in the future.
Robin Lee: As it transitions from pre-sale consultations to catalyzing more direct cells, our platform. Serving as a virtual cells person, Ernie Agent empowers advertisers to provide more personalized content and product introduction, thereby ensuring better pre-sale consultations.
Robin Lee: Early adopters from the education, legal and B2B sectors are pioneering Ernie Agent and have already seen a notable increase in effective cell fleets.
Robin Lee: We're proud to see that 16,000 advertisers already have their own Ernie Agents, which are now being distributed on the Baidu platform.
Robin Lee: Looking ahead, we believe Ernie Agent is poised to deepen its penetration and broaden its reach across more sectors, unlocking promising potential for our advertising benefits.
Robin Lee: And moving into intelligent driving, I mentioned earlier that Apollo Go has made remarkable breakthroughs in Wuhan, achieved 100% fully-dramanized operations in practically the entire municipality, reducing Q2 cost from vehicle by more than half compared to Q2 of last year.
Robin Lee: On top of that, Apollo Go has started scalable testing of the latest RG6 vehicles on open roads. These two factors are critical milestones in our strategy aimed at reaching UE breakeven in targeted cities.
Robin Lee: In addition to cost reduction, Apollo Go's operations continue to offer more expedient services.
Robin Lee: In Wuhan, its services is now available to 9 million people.
Robin Lee: The number of Apollo Go's pick-up points at the end of June has increased by over threefold from the previous quarter, enhancing accessibility and convenience for passengers.
Robin Lee: The deepening of our operational footprint and increased station density have propelled utilization rate for our vehicles, significantly driving up the daily rides per vehicle and the distance per ride.
Robin Lee: Nationwide, Apollo Go provided about 899,000 rides to the public in the second quarter, marking a 26% year-over-year increase.
Robin Lee: In July, the cumulative rides provided to the public surpassed 7 million.
Robin Lee: Even with these milestones, our share in the entire ride-hating service market is very small.
Robin Lee: It will take many years for us to reach a meaningful market share in China or elsewhere.
Robin Lee: Looking ahead, we are committed to providing increasingly affordable, convenient and safe travel for more passengers and drive long-term, sustainable growth.
Robin Lee: We will remain resolute in executing on our operational strategy, and at hosting efficiency and steering our intelligent driving bins towards profitability. In summary, by extrapolating the ongoing progress, from leveraging earnings to revolutionize product usage, and transform bins operations for our users, customers, and developers, and driven by a deep belief in technological innovation, we aim to create value for society and contribute to the greater good.
Rong Luo: Let me turn the call over to Rong to go through the financial results.
Rong Luo: Thank you, Robin.
Rong Luo: [inaudible] Nonggai Net income, attributable to Baidu was RMB 7.4 billion.
Rong Luo: Nonggai diluting earnings for ADS was RMB 21.02.
Rong Luo: Nonggai Net income, attributable to Baidu core was RMB 7.3 billion, and Nonggai Net margin for Baidu core was 27%.
Rong Luo: As of June 30, 2024, cash equivalence was[inaudible] with our pre-test, let's not open the clock to questions.
Operator: Ladies and gentlemen, we will now begin the question and answer session.
Operator: If you wish to ask a question, you will need to press star one on your telephone and wait for your name to be announced.
Operator: If you wish to cancel your request, please press star two.
Alex Yao: Your first question comes from Alex Yao with JP Morgan.
Alex Yao: Please ask your question.
Alex Yao: Thank you, management, for taking a question.
Alex Yao: I have a view on AI development.
Alex Yao: So first of all, how do you guys foresee the competition in AI evolving over the next two to three years?
Alex Yao: Will Baidu's current first mover advantage sustain, or will the technology gap among key players narrow over time?
Alex Yao: How do you guys view the current competitive landscape compared to, for example, Kimi, Doubao, Baba and Tenzin?
Alex Yao: In addition to technology leadership, what other factors are crucial to build a successful ecosystem in the AI industry?
Alex Yao: Thank you.
Robin Lee: Hi, Alex, this is Robin.
Robin Lee: Competition will be fierce over the next two to three years.
Robin Lee: Other may think whoever can make money from it will be able to survive.
Robin Lee: Last year, when China's AI industry was still at the so-called war of foundation models period, I predicted that the industry would go through consolidation, and only a few foundation models would survive.
Robin Lee: We are seeing a strong demand for our Gen AI products, and I'm proud that Ernie has continued to strengthen his market lead. And I'm confident we will maintain our leadership going forward. Our decade-long investment in AI provided us with the first mover advantage and technological leadership.
Robin Lee: You know, we were the first public company globally to launch a GPT-like model in March of last year. And since then, we keep leveraging our self-developed four-layer AI architecture for model upgrade.
Robin Lee: Last October, we launched the Ernie 4.0 that's China's first GPT-4 type model. And this June, we launched the Ernie 4.0 Turbo, which offers superior capabilities over 4.0.
Robin Lee: We will continue to draw out upgraded versions of our flagship foundation models.
Robin Lee: And our application-driven approach further expands our competitive advantage and set us apart from the others.
Robin Lee: We believe that foundation models hold little practical value without real applications built up on them, widely used by users and customers.
Robin Lee: So I mentioned earlier that currently Ernie handles about 600 million API calls and generates about 1 trillion tokens every day, which we believe is the highest in China.
Robin Lee: We have been heavily using Ernie to renovate our own products, like Baidu Search or Baidu Vancouver, into AI native applications on a large scale. And this approach has helped us gain valuable insights from users and continuously supported the upgrade of Ernie.
Robin Lee: And this experience allows us to create better solutions for our external customers when they use Ernie.
Robin Lee: At the same time, we have been working really hard to make Ernie increasingly accessible and affordable, so it can be used widely to address real world needs. We continue to lower the cost of model inference, expand our model portfolio, and develop two cases for model building and app building so that our customers, partners, developers can leverage the power of Ernie more efficiently and effectively.
Robin Lee: And this application-driven approach has helped us create a dynamic and expanding ecosystem.
Robin Lee: By the end of June, Petal Petal and Ernie developer community had grown to around 14.7 million.
Robin Lee: Meanwhile, thousands of thousands of enterprises are now engaged with us, highlighting the expanding influence and utility of our ecosystem.
Robin Lee: Looking ahead, we will be patient.
Robin Lee: We will continue to iterate on Ernie's capabilities to ensure we stay as a forefront of this new AI wave. By leveraging real world applications and feedback from our ecosystem partners, we gain valuable resources and insights to continuously refine our model.
Robin Lee: The synergy among our vibrant ecosystem, advanced model capabilities, and an application-driven approach will continue to strengthen our leadership in this rapidly evolving AI lens.
Robin Lee: Thank you.
Gary Yu: Your next question comes from Gary Yu with Morgan Stanley.
Gary Yu: Please ask your questions.
Gary Yu: Hi, thank you for the opportunity to ask questions.
Gary Yu: My questions are related to AI and search.
Gary Yu: So what kind of progress has been made in renovating traditional search using GNI and the last-language models?
Gary Yu: What kind of KBI is the company monitoring for AI and search, and have you have served any changes in user behavior, since implementing AI in the search business?
Gary Yu: I'm related to that.
Gary Yu: What is the role of applying for AI and search moving forward?
Gary Yu: And it puts specific KBI that the company aimed to achieve by the end of the year.
Gary Yu: For example, you know, certain percentage of search we saw generated by AI by the end of this year by in coming years.
Gary Yu: Thank you.
Robin Lee: Yeah, we are very committed to using cutting-edge AI technologies to transform our search capabilities and blend search with speed in one app.
Robin Lee: So KPI is related to Baidu app, well, override search-only KPI.
Robin Lee: In the second quarter, we accelerated the pace of renovating our search by leveraging Ernie.
Robin Lee: Our primary goal is to improve user experience, offer more capabilities, and boost user engagement on Baidu.
Robin Lee: As highlighted in my opening remarks, we made several important progress.
Robin Lee: First, we are providing smarter search results at an expanding pace.
Robin Lee: Currently, 18% of search results are provided by GNI.
Robin Lee: This results are more comprehensive and relevant than traditional search results.
Robin Lee: Offering information that was previously unavailable or scattered around different places.
Robin Lee: Furthermore, GNI enables infinite supply of constant on-demand and can effectively catch the decline of the open web system.
Robin Lee: Second, when GNI results can already meet user needs from the current query, we will recommend other contents or services just like a user would find in a typical new speed context. This will enhance user time spent and retention.
Robin Lee: Third, the new interactive function in Baidu search allows users to refine their questions and follow-off through multi-round conversations, reshaping how users express their needs and making the search experience more natural and efficient.
Robin Lee: Additionally, I would like to point out that we are rapidly rolling out Ernie agents with things search results to provide intelligent assistance to our users.
Robin Lee: By opening up our platform, developers can create a wide range of agents tailored to meet the unique needs of each and every user. Our vision is to create an ecosystem where many developers contribute to building a robust network of agents, collectively enriching the user experience and up-perform the most powerful frontier models, addressing a broad spectrum of challenges and opportunities.
Robin Lee: With all the efforts mentioned, the types of search queries on Baidu have become more diversified. Users are increasingly posing complex questions seeking comprehensive and in-depth answers from Baidu search. While this demands were once challenging to meet, they can now be efficiently addressed.
Robin Lee: On top of that, users are also using Baidu to create content such as writing and editing text and images.
Robin Lee: Generally, I is transforming Baidu search into a multi-functional creative platform offering a wider range of services to users.
Robin Lee: We are still in the early stages of an exciting yet uncharted opportunity.
Robin Lee: As we remain in the product, we are listening to feedback from users, creators, and partners.
Robin Lee: We are focused on long-term sustainability over short-term monetization.
Robin Lee: We will be patient and make sure that we set the right foundation for a new ecosystem and closely track user metrics.
Robin Lee: We believe this approach will deliver greater value to all our shareholders.
Robin Lee: Thank you.
Lincoln Kong: Your next question comes from Lincoln Kong with Goldman Sachs.
Lincoln Kong: Please ask your question.
Lincoln Kong: Thank you, manager, for taking my question.
Lincoln Kong: My question is about the cloud business.
Lincoln Kong: What are the drivers behind this accelerating growth of cloud revenue this year?
Lincoln Kong: How much of this incremental revenue will cloud this year's AI related and how does this contribute to the long-term AI cloud revenue growth?
Lincoln Kong: Also, what's the margin for the AI cloud compared to the traditional cloud business?
Lincoln Kong: How should we think about the intense pricing competition among different players in the market?
Lincoln Kong: [inaudible] So beginning a recognition across a wide range of sectors, from early adopters like internet companies and online education providers to the industries like automotive, financial services and utilities.
Lincoln Kong: Leading companies like Nasid Great, SFX Prize, and Guomin, Pension and Insurance, and also the rising AI startups, such as Shenzhou Technology, so they all intend to form in-depth partnership with us, leveraging our industry leading AI capabilities.
Lincoln Kong: So let's Robin just mention, a key revenue driver for AI cloud is the fast-growing Genie AI and Foundation model-related revenue. It has increased from about 5% of the total AI cloud revenue in a queue for last year to almost 9% this quarter.
Lincoln Kong: To make the diverse needs of our customers, we offer a wide range of products and solutions.
Lincoln Kong: So for those seeking model training and hosting services, we've provided the most cost-effective and powerful AI infrastructure.
Lincoln Kong: This quarter, we have observed a notable increase in spending on our public cloud from a GPU customers, reflecting their high satisfaction in trusting theirs.
Lincoln Kong: So this strong customer endorsement reinforces our confidence in continuing to gain market share.
Lincoln Kong: For the customers who are looking to harness the power of LNs, so API course, we offer the Ernie Finally, which includes some of the most powerful LNs available in the China market.
Lincoln Kong: Matt's platform also provides comprehensive two-kits for easy thanks-uning of cost and models and the development of applications tailored to us, especially these.
Lincoln Kong: We see the LLM and Genie AI market is still in its very early stages.
Lincoln Kong: We're educating the market and making AI accessible or key. On top of the model capabilities and ease of use, wide adoption also depends on affordability. So to encourage usage, we have made API course free for three late-week early models and cut price for flagship models. So this has led to a rapid increase in the early API course. Sorting just that, the total number of daily API course has searched from around 200 million in mid-May to over 600 million, which means over one trillion based tokens.
Lincoln Kong: We believe that the demand for Ernie's API course will continue to grow rapidly, given Ernie's superior and continuously improving efficiency and performance.
Lincoln Kong: For traditional cloud services, by cross-selling CPU cloud services to our existing GPU cloud customers. We have noticed a significant increase in the CPU cloud revenue from these customers.
Lincoln Kong: So overall, we are pretty confident that our AI cloud revenue grows.
Lincoln Kong: We have maintained strong momentum in the common quarters and capitalized on a significant long-term growth opportunities.
Lincoln Kong: So on the profit side, we are committed to sustainable and healthy growth, so our AI cloud revenue is healthier than ever in the consistently-generating non-gap operating profit. And we have been improving the margins.
Lincoln Kong: So regarding the price-compatient issue in your question, we all know that Geniei and our own market is still in this very area stage.
Lincoln Kong: And accessibility with superior price and performance is essential for growing the market.
Lincoln Kong: While we are working very hard to improve the capability of early models, we are also continuously lowering the inference cost.
Lincoln Kong: Considering the skill effect, as leading player in the market, I believe we will keep growing our much share.
Lincoln Kong: And also with the skill, I believe that the normalized margin for Geniei, really to the cloud business, should exceed that of traditional cloud business.
Lincoln Kong: Thank you.
Thomas Chong: Your next question comes from Thomas Chong with Jay Freys.
Thomas Chong: Please ask your question.
Thomas Chong: Hi.
Thomas Chong: Good evening.
Thomas Chong: Thanks, management, for taking my question.
Thomas Chong: Can management quantify the impact of the soft advertising revenue?
Thomas Chong: Distinguishing between week, month, or categorization impact from generative AI implementation, as well as the competition, particularly in search, separately?
Thomas Chong: What is the core advertising growth for the second half this year?
Thomas Chong: If we expect to see a turn around by year, and is it more driven by better macro or better monetization in AI, looking at 2025 and beyond what the sustainable growth rate.
Thomas Chong: Thank you.
Thomas Chong: Yeah.
Thomas Chong: [inaudible] The user time spend share change can be a proxy to the loss of our online advertising business.
Thomas Chong: And as I mentioned earlier, 18% of our search results contain genuine eye content, which is not monetized at this time.
Thomas Chong: This number will increase over time.
Thomas Chong: Of course, our aggressive raw-out of genuine eye features in the monetization system, partly offset the revenue sacrifice here.
Thomas Chong: Well, these are all the big challenges for the time being.
Thomas Chong: We strongly believe technological innovation will be the key for our future success. Specifically, we need to focus on what we can actively shape and influence, leveraging our strengths and resources to see new opportunities within our reach and a faster system for growth over time. So to this end, we are renovating our entire mobile ecosystem with genuine eye and earned.
Thomas Chong: For the advertising system, the renovations are already showing some results.
Thomas Chong: We talked about this during the previous quarters.
Thomas Chong: The incremental ad revenue generated by genuine eye and LLM continue to grow in Q2. This is mainly due to the improvement in the monetization system, which have improved ad conversion and efficiency.
Thomas Chong: And on top of that, we actively exploring opportunities for early agents to further enhance our advertising capabilities.
Thomas Chong: However, using AI to upgrade our existing advertising system is just a near-term step.
Thomas Chong: We believe that the true long-term value lies in the renovation of our consumer-facing products.
Thomas Chong: With over 20 years of experience in general search and our industry-leading capabilities in genuine eye and LLM, by doing uniquely positioned to innovate and transform general search into AI search.
Thomas Chong: This transformation is not just about enhancing the traditional search experience. It's about redefining what search can do for users.
Thomas Chong: Since Q2, we have accelerated the pace to renovate by the search using learning.
Thomas Chong: New features such as AI generated search without interactive features within search and early agents have been rolling out rapidly.
Thomas Chong: This renovation has high potential, but it's not something that will be completed in a short time. During this process, there may be a negative impact on revenue, but we are confident that the result will be a completely different new AI ecosystem, one that is much more versatile, productive, and user-friendly. So it may take some time before we see a noticeable revenue contribution from the AI enhancement in our search product, but we are confident that focusing on creating long-lasting user value rather than short-term revenue will ultimately create tremendous value and drive sustainable revenue growth.
Thomas Chong: Thank you.
Miranda Zhang: Your next question comes from Miranda Zhang with Bank of America Securities.
Miranda Zhang: Please ask your question.
Miranda Zhang: Thank you, management, for taking my questions.
Miranda Zhang: So my question is about the AI agent in the monetization.
Miranda Zhang: So can company explain how the AI agent is actually used in your user experience and monetization?
Miranda Zhang: How do you drive the adoption of this AI agent, especially among the SME advertisers, which have less data and less check capability?
Miranda Zhang: What kind of run-top pace do you expect by the end of this year for the AI agents?
Miranda Zhang: And also, can management give us more color on the recent development in the AI monetization, especially the progress from the CPC to the CPS model?
Miranda Zhang: What are the advertisers feedback on the effectiveness and the future budget allocation on the more than 10,000 early agents by now on our platform?
Robin Lee: The beauty of early agents is that anyone can create them, whether you are a user, a creator, a service provider, an advertiser, or a company. The barrier to entry is very low.
Robin Lee: It's really important to understand that the value of early agents extends far beyond just being a tool for monetization.
Robin Lee: It will significantly improve the user experience in our mobile ecosystem.
Robin Lee: A recent example is a lawyer agent created and distributed on our platform.
Robin Lee: A user has had a conflict with a law enforcement officer and started this confrontation with this agent.
Robin Lee: We had 288 runs of conversation and lasted for five hours.
Robin Lee: It's hard to imagine a real human lawyer would have the time and patience to talk to a potential client like that, but the benefits for the user, for the advertising law firm, and for our platform, are self-evident in this case.
Robin Lee: Today, early agent is still in its early stages and it's evolving quickly, but we're glad to see that advertiser by using early agent to get better selfie conversions than the traditional search.
Robin Lee: As I mentioned before, more and more advertisers, especially in fields like legal, education, and B2B, are starting to use Ernie Agent.
Robin Lee: This area is often required long decision-making times and personalized pre-sales confrontation.
Robin Lee: Ernie Agent, with its interactive features, is getting better and better at addressing this kind of needs.
Robin Lee: Plus, as Ernie Agent becomes smarter, it has the potential to drive direct sales and even more value for our advertisers. Therefore, moving us from a CPC model to a more effective CPS model.
Robin Lee: Ernie Agent, we believe Ernie Agent has the potential to become the new WebFact system in the AI era. Early stage Ernie Agent might not always improve the user experience right away, but as they start providing greater interactions, boosting engagement, and offering personalized services, monetization opportunities, will naturally follow.
Robin Lee: Thank you.
Alicia Yap: Your next question comes from Alicia Yap, West City Group.
Alicia Yap: Please ask your question.
Alicia Yap: Hi, thank you.
Alicia Yap: Good evening, management.
Alicia Yap: Thanks for taking my questions.
Alicia Yap: Have a question on the margin.
Alicia Yap: So Baidu Core continue to achieve the expansions in the operating margins in the second quarter.
Alicia Yap: Just wondering, is this expansion sustainable?
Alicia Yap: What is your outlook on the profit margin trend for Baidu Core in the entire year 2024?
Alicia Yap: Thank you.
Rong Luo: Thank you for a question.
Rong Luo: Let me hear a question.
Rong Luo: As you know, we don't give the guidance for quarter's hat.
Rong Luo: Baidu will try to give you some insights into our business priorities.
Rong Luo: What could they mean for our cost and expenses?
Rong Luo: For AI Cloud, we continue to face all the low margin businesses, so we can continue generating operating profit margins on a longer basis.
Rong Luo: Looking into the long-term, the normalized margins for the general AI-related cloud businesses should be higher than the Lexi cloud businesses.
Rong Luo: And this kind of revenue today is already not percent of our total cloud revenue.
Rong Luo: For intelligent driving and the other growth initiatives, we tend to reduce our losses, and particularly our intelligent driving businesses with the operational efficient gains, and try to reach your improvements in global tax savings.
Rong Luo: Overall, we believe that high-quality growth and the investments should be well balanced.
Rong Luo: For many years now, we have delivered a solid record of the top-line revenue growth and with a resolute cost-discipline and intend to continue building our futures on the basis.
Rong Luo: Sometimes, of course, there are investment runs at high-quality growth.
Rong Luo: We continue to operate our legacy such businesses in a challenging environment that pressures our revenue to tackle this kind of challenges.
Rong Luo: It's very important to focus on offering the uses, improving experiences, in particular a whole new AI base such experience for Baidu.
Xiaomeng Zhuang: Zhuang Li, Xiaomeng Zhuang Li, Xiaomeng Zhuang[inaudible] Xiaomeng Zhuang Li We're happy to see that more and more passengers are choosing a power goal, not just for testing rights, but as their go-to option for daily community.
Xiaomeng Zhuang: We'll handstand as our largest operational area with a fleet of around 400 vehicles offering 24 by seven service to 9 million residents there.
Xiaomeng Zhuang: Our focused on cost reduction through innovation has also made our services increasingly affordable.
Xiaomeng Zhuang: Since launching in Wuhan in Q3 of 2022, we've steadily increased the number of fully charging that 100% fully charged right-wing operation has been achieved in practically the entire city.
Xiaomeng Zhuang: This has a significant impact on costs. With our RT6 vehicles being tested at scale, we expect continued cost reduction once RT6 officially put into operation.
Xiaomeng Zhuang: So please note, our share of right-winging market in Wuhan is still very small and around 1% and the scaling will be a gradual process and it could take many years.
Xiaomeng Zhuang: The potential of China's right-winging market is enormous.
Xiaomeng Zhuang: Let's draw the tax that inherently offers more affordable services, it will naturally drive demand and open up great new opportunities.
Xiaomeng Zhuang: This is where innovation meets the vast scale of the market and creating a powerful synergy that will reshape the landscape of transportation and beyond.
Xiaomeng Zhuang: As Apollo goes scales up, it will offer an increasingly affordable, convenient and much affordable services to an expanding population.
Xiaomeng Zhuang: Also need to note that entering the robot taxi market requires deep believes in technology, long-term investment and strong capital reserves.
Xiaomeng Zhuang: Apollo goes stands out in this critical area, setting the benchmark for the industry.
Xiaomeng Zhuang: As for our question regarding future Benz model, currently all of our vehicles are owned by us and we bear the hardware, CapEx, and depreciation cost.
Xiaomeng Zhuang: But we are open to where with Benz models and partnerships.
Xiaomeng Zhuang: While our immediate focus is still achieving unique work even in case cities.
Xiaomeng Zhuang: In the long run, we aim to innovate as that light models like technology licensing and fleet operation to capture the huge market opportunities.
Xiaomeng Zhuang: And ultimately, benefiting society with better transportation solutions.
Operator: Thank you.
Operator: Ladies and gentlemen, that does conclude our conference for today.
Operator: Thank you for participating.
Operator: You may all disconnect.