Earnings Call Transcript - BIDU Q4 2023

Operator, Operator

Hello and thank you for standing by, for Baidu’s Fourth Quarter and Fiscal Year 2023 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.

Juan Lin, Director of Investor Relations

Hello everyone and welcome to Baidu’s fourth quarter and fiscal year 2023 earnings conference call. Baidu’s earnings release was distributed earlier today and you can find a copy on our website as well as on our newswire services. 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 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 Harbor Provision of the U.S. Private Securities Litigation Reform Act of 1995.

Robin Li, CEO

Hi. And we concluded 2023 on a solid note. Baidu’s total revenue for the fiscal year increased about 8% year-over-year, and the non-GAAP operating margin expanded from 22% to 24%. This improvement highlighted the resilience of our core business, which serves as a strong foundation for our ventures in ERNIE and ERNIE Bot. 2023 marked a year of Gen AI and foundation models, a massive technology shift that will re-engineer processes and bring about a real renaissance in every sector of the economy. Baidu is well prepared to benefit greatly from this groundbreaking economic revolution. We progressed from discussing the opportunities of Gen AI and foundation models to actively implementing high ERNIE and ERNIE Bot at scale. As the front-runner in AI, Baidu has probably become the first public company globally to launch a GPT model, with our EP 4.0 standing high as the most powerful foundation model in China. ERNIE continues to gain market recognition, as evidenced by ERNIE API calls from multiple well-known companies. Notably, Samsung uses ERNIE API on its Galaxy S24 5G sales, Honor uses ERNIE API in its Magic 8.0, and Autohome utilizes ERNIE API to power multiple AITC apps. The partnerships further solidify our leadership in Gen AI. Furthermore, we continued improving the efficiency of ERNIE. For example, the inference cost of EB 3.5 is only about 1% of the March 2023 version. With lower inference costs, ERNIE will become increasingly accessible to users and enterprises. We are able to do that primarily because of our unique four-layer AI architecture and our strong ability in end-to-end optimization. Additionally, we are offering our RMs in smaller sizes. These small models can balance performance and efficiency, better serving customer needs. Since Q2 2023, we have actively utilized ERNIE to revolutionize our products and services, creating AI-native experiences. We believe real applications are essential to unleashing the full business potential of ERNIE and ERNIE Bot. Recently, we began to generate incremental revenues from ERNIE and ERNIE Bot. In the fourth quarter, we earned several hundred million RMB primarily from ad technology improvement and helped enterprises build their own models. I'll provide a more detailed explanation in the business review section. Looking into 2024, we believe this incremental revenue will multiply to several billion RMB primarily from advertising and AI cloud building. While we are beginning to commercialize Gen AI and foundation models, we see enormous monetization potential in this groundbreaking technology. We envision ERNIE as the future foundation system, serving as the backbone for millions of AI-native applications developed by third parties and Baidu. This paradigm will enable us to create an ecosystem around ERNIE, which opens up various revenue sources. Now, the key highlights in Baidu’s native AI applications. Equipped with AI copilot, Baidu Wenku has become a platform for users to create content in a wide range of formats, helping them express their ideas. In addition to summarizing and creating content, users particularly appreciate Baidu Wenku’s features for assisting them in automatically turning their inspiration into PowerPoint presentations. We have been consistently improving and enriching its AI features. For example, we recently introduced new features, such as Mind Maps for organizing information. We are reconstructing Baidu search using ERNIE. Last quarter, we talked about how we used ERNIE to generate search results, improving the on-site experience. In Q4, Baidu search introduced a news feed-like information alongside the generated search results at scale to offer users useful and relevant content to explore. In Q4, we also began facilitating certain search queries with multi-round conversations to delve deeper into understanding user intent. Overall, Gen AI-enabled search complements traditional search, expanding the total addressable market (TAM) of Baidu search by serving a wider range of information and answering content creation-related questions. We are actively encouraging SMEs in service industries to build AI chatbots as new landing pages for Baidu search. AI chatbots enable them to better serve potential buyers directed to them by us, driving transactions on our platform. At the same time, our AI chatbot can help merchants increase productivity. Currently, about 4,000 merchants are using our AI chatbots, particularly in healthcare, education, travel, legal, B2B and auto, where users often interact with merchants to learn about their offerings before making purchasing decisions. We believe that Gen AI-enabled search will help increase user intention on Baidu, reinforcing our position as the primary entry point for users seeking information and services. We are also expanding the range of ERNIE by enabling enterprises to easily develop AI-native applications with ERNIE. In December, we introduced an app builder, a comprehensive set of tools on our public cloud to support enterprises in swiftly building AI applications from the ground up. This tool enables enterprises to incorporate capabilities, such as organizing unread messages for prioritization, creating drafts, summarizing meeting notes, and facilitating Gen AI-enabled data analysis. Additionally, we have included several ready-made templates, such as AI agent, to enable enterprises to promptly create applications for validating their ideas. Such progress is largely based on our own pioneering experience in developing AI-native applications. In addition to the tools for app development, enterprises partner with us because we have the most powerful and most efficient foundation models, along with the most cost-effective cloud platform for running applications. In December, about 26,000 enterprises were actively using ERNIE through our API on a monthly basis, representing an increase of 150% quarter-over-quarter. ERNIE is now handling more than 50 million queries every day. That's up 190% quarter-over-quarter, demonstrating a significant rise in third-party activity. As the number of AI-native applications hosted on our public cloud continues to rise, we are poised to accelerate the enhancement of our model performance by incorporating feedback from our users and customers. This will further widen the gap between ERNIE and our domestic peers. Moreover, as more and more AI-native applications on ERNIE become popular, ERNIE will likewise continue to thrive, supporting a robust ecosystem around it and steering our multi-year growth forward. In conclusion, we find ourselves amidst tremendous opportunity in Gen AI and foundation models. With ERNIE and ERNIE Bot, we have started generating incremental revenue. In 2024, we expect AI revenues to become more meaningful while our core business will remain resilient. Moreover, we are preparing to serve the next wave in the development of ERNIE. In the future, we believe that there will be millions of AI-native applications powered by ERNIE, with the majority of them developed by our cloud customers. The expanding ERNIE ecosystem will then unlock numerous monetization potentials for Baidu. Now let me recap the key highlights for each business for the fourth quarter of 2023. The mobile ecosystem exhibited solid performance across revenue, margin, and cash flow. In the fourth quarter, Baidu’s core online marketing revenue increased by 6% year-over-year, driven by verticals in travel, healthcare, business services, and others. In Q4, we generated several hundred million RMB in incremental ad revenue due to improvements in advertising technology. We continue to use ERNIE to enhance our monetization system, including targeting capability and bidding system in the quarter. Starting from January, our monetization system has been able to generate real-time text-based ads for search queries, demonstrating significant technology improvements in both targeting and ad creation. As of today, our new marketing platform has attracted about 10,000 advertisers. An advertiser in the medical aesthetics industry began to adopt our new marketing platform in Q4. With the AI copilot, if the advertiser can articulate its requirements comprehensively through native language and multi-round conversations for further clarification and explanation, the conversational experience has assisted in building optimized search and feed campaigns by generating relevant ad content. Our platform has also helped them reach highly targeted audiences while dynamically allocating budget to drive conversions. The advertiser achieved a 22% increase in conversion rate and a 5% reduction in cost for sales lead acquisition after using this enhanced platform and capabilities. Currently, only a small portion of our entire advertiser base has adopted this new marketing platform, which means there is a significant opportunity for future growth. All these initiatives have helped advertisers improve their conversion rates, leading to an increase in their ad budget on our platform in the quarter. We expect continued growth in ad revenue from Gen AI and foundation models in 2024 and beyond. While we enhanced our ad technologies, we also introduced AI chatbots for brands, an innovative ad product built on top of ERNIE Bot in Q4. AI chatbots have further enriched our product portfolio. They can enhance user engagement and customer surveys while also capturing customer attention and driving customer demand. Since last October, China Feihe, a local infant formula company, has adopted an AI chatbot to promote its brand. The AI chatbot has been enabled for Facebook to have multi-round interactions with our customers, increasing brand recognition and providing valuable insights into potential consumers. These profound insights have also helped the player Facebook understand customer views of the brand and products, leading to an enhanced marketing strategy. The AI chatbot for brands has garnered significant interest from brand advertisers, and we expect more interest to embrace the AI chatbot in the future. As I previously mentioned, we're using ERNIE to build our apps. With continuous enhancement of our AI-native product offerings, we believe more users will come to the Baidu platform. This, in turn, will assist us in gaining more market share in user engagement and online advertising revenues. Looking into 2024, the mobile ecosystem should continue to generate steady profits and cash flow. AI Cloud revenue grew by 11% year-over-year to RMB 5.7 billion and continued to improve profitability in the fourth quarter. Revenue from Gen AI and foundation models represents 4.8% of our AI Cloud revenue in Q4. The increasing demand for model building played a significant role in this accelerated revenue growth, along with increasing distributions from inference. We have seen a growing number of enterprises, particularly tech companies, turning to our public cloud to build their models. Additionally, the AI cloud revenue generated by Baidu Core and other business groups, such as the Mobile Ecosystem Group and the Intelligent Driving Group was about RMB 2.7 billion in Q4. Within the Q4 internal cloud revenue, Gen AI and foundation models accounted for about 14%. On a combined basis, the total internal and external AI Cloud revenue was RMB 8.4 billion in Q4, with Gen AI and foundation models contributing around RMB 656 million. Gen AI and LLM have become pivotal considerations for many enterprises, driving a shift in the cloud industry from general-purpose computing to AI computing. This evolution is reshaping the competitive dynamics within the cloud industry, strengthening our lead in AI and expanding our total addressable market. Enterprises choose us, thanks to our position as the most powerful and cost-effective AI infrastructure for both model training and inference in China. Thanks to our unique four-layer AI infrastructure driving consistent efficiency gains and years of experience in ERNIE. Last quarter, we explained that 98% of LLM training time on our AI infrastructure was valid. Moreover, our GPU networking is running at a 95% utilization for training LLMs. Both of these metrics have set industry benchmarks. Within our Maps, we offer a model builder and app builder, which are two sets of tools that help enterprises effortlessly build models and develop apps. As of now, enterprises have built about 10,000 models on our Maps. Since its inception, App Builder has facilitated the creation of thousands of AI-native apps. As more and more applications are being built on our Maps, we will have greater revenue potential going forward. Looking into 2024, AI Cloud should maintain strong growth in revenue and generate profit at a non-GAAP operating level. Our intelligent driving business continued to focus on achieving new breakeven for Apollo Go. In Wuhan, Apollo Go's largest operation, about 45% of our orders were provided by fully driverless vehicles in Q4. This metric surpassed 50% in January. The increase is due to our intensified operations during peak hours in areas with complex traffic conditions and further expanding our operating area in the past few months. This development resulted from our ongoing efforts to improve technology through safely operating Apollo Go on public roads. In China, Apollo Go provided about 839,000 rides in the public in Q4, marking a 49% year-over-year increase. In early January, the cumulative rides offered by Apollo Go exceeded 5 million. The substantial data collected from operations will further help us enhance the efficiency of safe operations. Looking into 2024, we will remain focused on getting closer to Apollo Go's unit economic breakeven target and managing our costs and expenses to reduce losses in intelligent driving. Upon reaching unit economic breakeven, we plan to swiftly replicate our successful operations in Wuhan to other regions. In summary, we are facing tremendous opportunities in Gen AI and foundation models. We will continue to invest in these opportunities. At the same time, we will strive to optimize our cost and expense structure for each business line to improve operational efficiency. With that, let me turn the call over to Rong to go through the financial results.

Rong Luo, CFO

Thank you, Robin. Now let me walk through the details of our fourth-quarter and full-year 2023 financial results. We closed 2023 with solid financial results. Total revenues in the fourth quarter were RMB 35 billion, increasing 6% year-over-year. Total revenues for the full year 2023 were RMB 134.6 billion, increasing 9% year-over-year. Baidu Core's Q4 revenue was RMB 27.5 billion, increasing 7% year-over-year. In 2023, Baidu Core generated RMB 103.5 billion in revenue, increasing 8% year-over-year. Baidu Core's online marketing revenue increased 6% year-over-year to RMB 19.2 billion in Q4, accounting for 7% of Baidu Core's total. Baidu Core online marketing revenue was up 8% year-over-year in 2023. Baidu Core's non-online marketing revenue was RMB 8.3 billion, up 9% year-over-year. For the full year 2023, our online marketing business increased 9% year-over-year. The increase in our online marketing business was mainly driven by the AI Cloud business. Revenue from IT was RMB 7.7 billion in Q4, increasing 2% year-over-year. Revenue from IT was RMB 31.9 billion in 2023, increasing 10% year-over-year. Cost of revenue was RMB 17.4 billion in Q4, increasing 3% year-over-year, primarily due to an increase in costs related to AI Cloud business, partially offset by a decrease in quantum costs. Cost of revenue was RMB 65 billion in 2023, increasing 2% year-over-year primarily due to an increase in traffic acquisition costs, partially offset by a decrease in content costs and the costs related to AI Cloud business. Operating expenses were RMB 12.1 billion in Q4, increasing 5% year-over-year, primarily due to an increase in server depreciation expenses and server custody fees, which support Gen AI research and development improvements and channel spending and promotional marketing expenses. Operating expenses were RMB 47.7 billion in 2023, increasing 9% year-over-year, primarily due to an increase in channel spending and promotional marketing expenses and server depreciation expenses and server custody fees, which support Gen AI research and development improvements. Operating income was RMB 5.4 billion in Q4. Baidu Core's operating income was RMB 4.7 billion, and Baidu Core's operating margin was 17% in Q4. Operating income was RMB 21.9 billion in 2023. Baidu Core's operating income was RMB 18.8 billion, and Baidu Core's operating margin was 18% in 2023. Non-GAAP operating income was RMB 7.1 billion in Q4. Non-GAAP Baidu Core operating income was RMB 6.2 billion, and non-GAAP Baidu Core operating margin was 23% in Q4. Non-GAAP operating income was RMB 28.4 billion in 2023. Non-GAAP Baidu Core operating income was RMB 24.7 billion, and non-GAAP Baidu Core operating margin was 24% in 2023. In Q4, total other loss net was RMB 2.5 billion compared to total other income net of RMB 1.8 billion for the same period last year, mainly due to a peak of losses from equity method investments as a result of a modification of certain terms of underlying preferred shares. Income tax benefit was RMB 96 million. In 2023, total other income net was RMB 3.3 billion compared to total other loss net of RMB 5.8 billion last year, mainly due to a fair value gain of RMB 198 million from our long-term investments this year compared to a fair value loss of RMB 3.9 billion last year and a decrease of RMB 2.2 billion in impairment of long-term investments. Income tax expenses were RMB 3.6 billion. In Q4, net income attributable to Baidu was RMB 2.6 billion, and diluted earnings per ADS was RMB 6.77. Net income attributable to Baidu Core was RMB 2.4 billion, and net margin for Baidu Core was 9%. Non-GAAP net income attributable to Baidu was RMB 7.8 billion, non-GAAP diluted earnings per ADS was RMB 21.86. Non-GAAP net income attributable to Baidu Core was RMB 7.5 billion, and non-GAAP net margin for Baidu Core was 27%. In 2023, net income attributable to Baidu was RMB 20.3 billion, and diluted earnings per ADS were RMB 55.08. Net income attributable to Baidu Core was RMB 19.4 billion, and net margin for Baidu Core was 19%. Non-GAAP net income attributable to Baidu was RMB 28.7 billion, non-GAAP diluted earnings per ADS was RMB 80.85. Non-GAAP net income attributable to Baidu Core was RMB 27.4 billion, and non-GAAP net margin for Baidu Core was 26%. As of December 31, 2023, cash, cash equivalents, restricted cash, and short-term investments were RMB 205.4 billion, and cash, cash equivalents, restricted cash, and short-term investments, excluding IT, were RMB 200 billion. Free cash flow was RMB 25.4 billion, and free cash flow, excluding IT, was RMB 22.1 billion in 2023. Baidu Core had approximately 35,000 employees as of December 31, 2023. With that, operator, let's now open the call to questions.

Operator, Operator

Thank you. We will now begin the question-and-answer session. Our first question today will come from Alicia Yap of Citi. Please go ahead.

Alicia Yap, Analyst

Hi, thank you. Good evening, management. Thanks for taking my questions. My question is on the outlook. How does management think about the macroeconomic landscape in China for 2024? What's management view on 2024 growth outlook for Baidu as a whole? And also, what is the percentage of total revenues of Baidu that will be contributed from AI-related revenue in 2024? Thank you.

Robin Li, CEO

Hi, Alicia, this is Robin. Before we dive into the outlook, let's take a quick look back at last year. We had a very challenging macroeconomic environment, but our business demonstrated very solid performance. We invested aggressively in Gen AI, but our non-GAAP operating margin expanded year-over-year, and revenue experienced solid growth. More importantly, we began to generate incremental revenue from Gen AI and foundation models. For this year, we have noticed that both the central and local governments are putting efforts to grow the economy. During the eight-day Chinese New Year holiday, we saw a growth in consumption, particularly in the travel sector. But we are still operating in a macro environment with a lot of uncertainty. We are closely monitoring significant economic stimulus plans, which we think are essential for achieving this year's goals. That being said, we are facing a lot of opportunities. Our core business remains solid, and the incremental revenue from Gen AI and foundation models will increase to several billion RMB in 2024, contributing to the growth of our total revenue. More specifically, thanks to our leading position in LLM and Gen AI, enterprises are increasingly building models and developing apps on Baidu Cloud. Our mobile ecosystem has already accumulated a huge user base, and we keep renovating our products and enhancing our monetization capabilities through AI innovation. Combining Cloud and Mobile, we believe we will be able to sustain our long-term growth, which we think will be faster than China's GDP growth. Thank you.

Alicia Yap, Analyst

Thanks.

Operator, Operator

Okay. And our next question today will come from Alex Yao of JPMorgan. Hi, Alex, your line is open. Are you muted?

Alex Yao, Analyst

Yeah. Sorry, I was muted. Good evening, management. Thank you for taking my questions. I have a couple of questions on cost structure and the margin trends. First of all, how much more room do we see in cost cutting and optimization, and how should we look at AI-related investments? In the past, you guys discussed that there will be a lag between chip investments and AI revenue contribution? How should we look at the margin trend into 2024 if you intend to expand?

Rong Luo, CFO

Alex, thank you so much for your questions. This is Rong. I think alongside the investments in our Gen AI businesses, we are sure that there is still room to manage the costs and expenses in our legacy businesses. As we look into 2024, we will continue focusing on our core businesses, and we will also be ready to reduce our resource allocations to non-strategic businesses. Additionally, we are continuously enhancing our overall organizational efficiencies by removing layers, simplifying executions, and flattening the organization structures. So heading into this year, we are very committed to the ongoing optimization of our operations, ensuring we have a more productive HR team. With all these measures in place, we aim to keep Baidu Core's earnings solid, with our mobile ecosystem continuing to deliver strong margins and generate very strong cash flows while the AI Cloud business continues its profitability. We have managed to maintain solid operating profit margins despite our investments in AI. If you remember from last year, we started to invest in Generative AI and large language models. This investment is mainly reflected in our capital expenditures, primarily related to purchasing chips and servers for model training and services. As you know, the CapEx will be depreciated over several years, so despite a 58% year-over-year increase in Baidu's CapEx in 2023, our non-GAAP operating profit margins still saw a 2 percentage point growth on a year-over-year basis. Looking forward, while it is inevitable that we will make new investments during the process of developing our new AI business, these investments are not expected to significantly affect our margins or profits. During the early stage of market development, we will not overly prioritize margins for our AI Cloud business, as we believe that in the long run, this business is expected to yield much better margins. Additionally, there may be some promotional activities for the AI-native 2C products, but we will carefully manage and closely monitor to balance investments and growth. We are happy to see that our efforts in investing in new initiatives have begun to yield early results. As Robin mentioned earlier, the incremental revenue generated from ad tech improvements reached several hundred million RMB in Q4, and the incremental AI Cloud revenue generated from Gen AI and foundation models also contributed 4.8% of the total AI Cloud revenue. I think all these promising top-line achievements have strengthened our confidence in our business strategy. So going forward, we will remain steadfast in our commitment to developing Gen AI and large language model use. Thank you, Alex.

Operator, Operator

Our next question today will come from Gary Yu of Morgan Stanley. Please go ahead.

Gary Yu, Analyst

Hi, good evening, and thank you for taking my question. I have a question regarding AI contribution. So for AI-related revenue contribution, is there a way you can quantify or prove that the ad revenues Baidu will be generating is purely incremental contribution from AIGC and not cannibalizing from your existing search business? And if AI is purely incremental, should we be expecting faster than average growth? Additionally, excluding AI, how should we think about the core growth rate for 2024? Thank you.

Robin Li, CEO

Hi, Gary, this is Robin. We are the largest search engine in China, with close to 700 million monthly active users. We have established a robust brand presence among the Chinese Internet and mobile users. They rely on us for comprehensive and reliable information. So we have a strong and stable base of revenue and profits. However, we are also very sensitive to macroeconomic conditions because our advertising business has a broad coverage of different verticals. I mentioned earlier that there are still uncertainties regarding the macro environment. However, Gen AI and LLM are unlocking new opportunities at both the monetization front and the user engagement front. I think it's easier to quantify the incremental revenue on the monetization front. I mentioned earlier that Gen AI has already helped increase advertising ECPM, and our upgraded monetization system has allowed us to improve targeting capabilities, thereby generating and presenting more relevant ads. We earned several hundred million in Q4 from this kind of initiative, and the incremental revenue will grow to several billion this year. It's harder to quantify the user engagement part. Gen AI is helping us improve user experience. We have seen initial outcomes in search and our Wenku platform, and we will continue introducing new features going forward. These initiatives will help us improve user mind share and time spent over time and bring us even larger potential. So I think the purely incremental revenue will come from both the monetization and the user engagement sides. Thank you.

Operator, Operator

Our next question today will come from Lincoln Kong of Goldman Sachs. Please go ahead.

Lincoln Kong, Analyst

Hi, thank you, management, for taking my questions. My question is regarding the Cloud business. How should we look at the incremental revenue growth driven by Generative AI? What is the product mix for Generative AI Cloud? And what exactly are the offerings that are primary growth drivers here? Furthermore, when we're looking at 2024, how should we expect the overall AI Cloud revenue growth this year, and what will be the margin trend this year as well? Thank you.

Rong Luo, CFO

Thank you for the question. This is Rong. As Robin just mentioned, the total revenue from our Gen AI and foundation model-related businesses, including both internal and external revenue, already reached RMB 656 million in Q4, and this number should grow to several billion RMB for the full year 2024. We have seen increasing interest from enterprises in using Gen AI and LLM to develop new applications and features. To achieve this goal, they are actively building models to power their products and solutions, which is how we generate the majority of revenue from external customers. Meanwhile, we are seeing significant increases in model inferencing revenue from external customers. So while revenue from inferencing is still small, we believe that over the long term, it will become a sustainable revenue driver. We think revenue generated from internal customers is also quite important because a significant portion of such revenue is from model inferencing. Baidu is the first company to use Gen AI and LLM to reconstruct all of our businesses and products. As the number of products and features powered by Gen AI and LLM continues to increase, ERNIE API costs from internal customers have been rapidly increasing and have reached a significant magnitude. As such development has proven that ERNIE and ERNIE Bot can effectively enhance productivity and efficiency in real-world applications. Regarding our product offerings, we have the most powerful AI infrastructure for model training and inferencing in China. This infrastructure helps our customers build and run models cost-effectively. Additionally, as Robin mentioned earlier, our models offer various options and a full set of tools in terms of model builder and app builder for model building and application development. In addition to that, we have developed our own AI-native solutions, such as Generative Business Intelligence (GBI), that helps enterprises increase productivity and efficiency. Beyond the incremental opportunity related to AI, Gen AI and foundation models bring new opportunities to our legacy cloud businesses. We continue to win customers and projects for CPU-centric cloud because we are highly recognized for our AI infrastructure and CPU-centric cloud offerings, especially in the Internet, tech, and education sectors. Gen AI and foundation models have allowed us to build AI solutions for our customers more efficiently than before, facilitating digital and intelligent transformations for traditional industries. Both of these factors are driving growth for our cloud revenue. Overall, we should see cloud revenue growth accelerate in 2024 compared to last year. Additionally, we are quite confident in maintaining profitability for our AI Cloud. For Enterprise Cloud, we should be able to consistently improve gross margins for the legacy cloud businesses. As for Gen AI and LLM businesses, the market is still in a very early stage of development. Therefore, we should hold a dynamic pricing strategy to quickly educate the market and expand our penetration into more enterprise customers. We believe that over the long term, the new business will yield higher normalized margins than traditional cloud businesses.

Operator, Operator

Our next question today will come from Thomas Chong of Jefferies. Please go ahead.

Thomas Chong, Analyst

Hi, good evening. Thanks, management for taking my questions. May I ask about the pace for our AI 2C product development? How has the traffic growth been? Are there any key metrics you can share on the new generative search? How does AI benefit search traffic and time spent? And when should we expect the exposure of traffic or super app? Thank you.

Robin Li, CEO

Yes, we are reconstructing all of our 2C products using generative AI. I think Gen AI and foundation models are making all of our products more powerful. Over the past few months, we have made significant strides in this area and the initial user feedback has been encouraging. For search, the introduction of Gen AI has enabled Baidu to answer a broader range of questions, including more complex, open-ended, and comparative queries. With ERNIE Bot, we can provide direct and clear answers in a more interactive way. During the past few months, instead of just lending some content and providing links, more and more search results are generated by ERNIE Bot. As a result, users are engaging with Baidu more frequently and asking new questions. For example, more users are coming to Baidu for content creation, be it text or images. During the Chinese New Year holiday, Baidu helped users create New Year greeting messages and generate personalized e-cards for their loved ones. This is not a typical use case for a search engine, but we see a significant number of users relying on Baidu for this kind of need. Going forward, we will increasingly use ERNIE Bot to generate answers for search queries and use multi-round conversations to clarify user intent so that complicated needs can be addressed through natural language. While these initiatives have resulted in an enhanced search experience, we are still in the early stages of using ERNIE Bot to reconstruct Baidu search. We will continue to test and iterate Gen AI-enabled features based on user feedback, following our typical playbook for testing and fine-tuning new experiences before they are ready for larger-scale rollout. Overall, we believe Gen AI will complement traditional search, ultimately increasing user retention, engagement, and time spent on Baidu. In addition to search, ERNIE Bot acting as a copilot in Wenku has transformed from an application for users to find tablets and documents to a one-stop shop for users to create content in a wide variety of formats. Year-to-date, approximately 18% of new paying users were attracted by the Gen AI features for Wenku. I want to emphasize that we are in the early stages of using ERNIE Bot to reconstruct our apps and build new ones. At the same time, we are attracting and helping enterprises build apps on ERNIE. We believe that the success of ERNIE depends on its wide and active adoption, whether through Baidu apps or third-party apps. Thank you.

Operator, Operator

Our next question today will come from James Lee of Mizuho. Please go ahead.

James Lee, Analyst

Great. Thanks for taking my questions. Can you talk about ERNIE's technology roadmap for 2024? Does it include multimodal features, maybe similar to Sora or launching an AI store or potentially opening an AI engine? Are there any milestones or key metrics you can speak to? The second part of that question is on the cost of running Gen AI; how should we think about managing inferencing costs going forward? Are there any additional levers to optimize this process?

Robin Li, CEO

Yes, the chips we have on hand should enable us to advance ERNIE Bot to the next level. As I mentioned earlier, we will take an application-driven approach to enhance ERNIE, based on our users and customers telling us where we should improve and adjust our models. This could involve building multi-modal capabilities, enhancing reliability, and so on. It's important to note that we are focusing on using ERNIE to bring real value to users and customers rather than simply achieving high rankings in downloads and research publications. Price is also a very important issue; making high-performance foundation models affordable is critical for large-scale operations. I mentioned earlier that we have been continuously reducing model inference costs, and currently, the inference cost of ERNIE Bot 3.5 is about 1% of the March 2023 version. By doing so, more enterprises are increasingly willing to test, develop, and iterate their applications on ERNIE. We understand that for many customers, they tend to balance efficiency with cost and speed, so we also offer smaller language models and help our customers leverage MOE (Mixture of Experts) for optimal performance. With our end-to-end approach, we believe there is still ample room to reduce the costs of our most powerful models and make them increasingly affordable to our customers. This will help further drive the adoption of ERNIE. Internally, we are closely monitoring the number of apps developed on ERNIE. As I mentioned previously, ERNIE is now handling over 50 million queries per day. Currently, ERNIE API costs from internal applications are still larger than external app costs. The costs for ERNIE in different sizes from external apps have been increasing rapidly. We are just at the beginning of this journey. ERNIE will only become more powerful, smarter, and more useful as more end users utilize it, whether through Baidu apps or third-party applications. This will enable us to cultivate an ecosystem around ERNIE. As these apps and models are actively used by end users, they will also generate significant inferencing revenue for us. Thank you.

Operator, Operator

Our next question today will come from Charlene Liu of HSBC. Please go ahead.

Charlene Liu, Analyst

Thank you. Good evening, management, and thank you again for the opportunity. I have a question related to ERNIE. How does enterprise adoption compare to its peers? Can you please kindly share with us the latest number of enterprises that are using ERNIE to build models and applications and help us understand how that has gone versus last quarter and what the underlying drivers may be? Lastly, can you help us also understand whether we can assume that enterprises who apply the ERNIE API integration will very unlikely be using other elements?

Rong Luo, CFO

Great question, Charlene. So as Robin just mentioned, we have about 26,000 enterprises of different sizes, spreading over different verticals, using our ERNIE API, which increased 150% quarter-over-quarter. The ERNIE API costs have exceeded 50 million on a daily basis, and we believe no one else in China has gained so many customers or received such high volumes of API requests. The enterprises choose our platform primarily for the following reasons. Firstly, we have the most cost-efficient AI infrastructure for model training and inferencing in China, primarily due to our strong ability in end-to-end optimization. Moreover, Gen AI and large language models are reshaping the competitive landscape of China's public cloud, enhancing our competitive advantage. Our strong ability in managing an extensive GPU-centric cloud with very high GPU utilization has continuously enhanced our AI infrastructure, allowing us to help enterprises build and run their models and develop AI-native applications at a low cost on our cloud. Secondly, the ERNIE family of models has attracted many customers to our cloud. Over the past few months, we have consistently enhanced ERNIE’s performance, receiving positive feedback from customers. We also offer ERNIE models in different sizes to accommodate customers' needs regarding cost structures. Thirdly, we were the first company in China to launch a model-as-a-service offering, which is a one-stop shop for LLM and AI-native application development. ERNIE Bot makes it easier for enterprises to use LLMs. We also provide toolkits to help enterprises easily tune or fine-tune their models and develop applications on our cloud. These toolkits allow customers to create their models cost-effectively by incorporating their proprietary data and use ERNIE API to power their applications directly. We also assist them in supporting different product features using different models, adopting the MOE approach in app development. As a result, enterprises can focus on identifying customer pain points rather than expanding their efforts on programming. All these initiatives have helped us establish a first-mover advantage in Gen AI and LLMs. As more customers use our cloud platform to develop AI-native applications aimed at attracting users, substantial user and customer insights will be generated and accumulated. These insights will help us refine the toolkits. As our tools become increasingly customer-friendly and help enterprises effortlessly finetune models and create apps, they will be more inclined to stay with us. Additionally, at the current stage of employing large language models, it is crucial for customers to create suitable prompts for their chosen models. Since they have to invest considerable effort in building and accumulating their best prompts for using large language models, it becomes less sensible for them to switch to another model as they will have to re-establish their prompt portfolio. Thus, with increasing adoption and active usage of our platform, customer satisfaction and switching costs will help us increase customer retention.

Charlene Liu, Analyst

Thank you.

Operator, Operator

Our next question today will come from Miranda Zhuang of Bank of America Securities. Please go ahead.

Miranda Zhuang, Analyst

Good evening. Thank you for taking my questions, which are about the AI chip. I'm wondering what's the impact on your AI development after the recent further chip restrictions from the U.S.? Is there any update on the alternative chips? Given the chip concern, how is Baidu developing the AI model product and monetization differently compared to overseas peers? What can be achieved and what may become difficult? What will the company do to keep up with overseas peers in the next few years? Thank you.

Robin Li, CEO

In the near term, the impact is minimal for our model development or product inventions or monetization. As I mentioned last quarter, we already have the most powerful foundation model in China, and our AI chip reserve enables us to continue enhancing ERNIE for the next one or two years. For model inference, it requires less powerful chips. Our reserve and the chips available on the market are sufficient for us to power many AI-native applications for end users and customers. In the long run, we may not have access to the most cutting-edge GPUs. However, with the most efficient homegrown software stack, the user experience will not be compromised. There is ample room for innovation in the application layer, model layer, and framework layer. Our end-to-end self-developed four-layer AI architecture, along with our strong R&D team, will support us in using less advanced chips for efficient model training and inference. This provides Baidu with a unique competitive advantage over our domestic peers. For enterprises and developers, building applications on ERNIE will be the best and most efficient way to embrace AI. Thank you.

Operator, Operator

Our next question today will come from Ken Fong of UBS. Please go ahead.

Unidentified Analyst, Analyst

Good evening, management. This is Ken Fong from UBS. Thank you for taking my question. Recently, we have witnessed significant developments in text-to-video and video generation technology. How do you see this technology affecting the broader AI industry in China, and what implications does it have for ERNIE? Could you detail your strategic plan for ERNIE going forward? Additionally, how is ERNIE performing in text generation, as well as text-to-image and text-to-video generation tasks? What improvements do you anticipate in these areas?

Robin Li, CEO

This is Robin. First of all, multi-modal integration, such as text-to-audio and video, is an important direction for future foundation model development. It is a must-have for AGI. Baidu has already invested in this area and will continue to do so in the future. Secondly, if we look at the development of foundation models, the market for large language models is huge and still at very early stages. Even the most powerful language models in the world are still not sufficient for many applications. There is still plenty of room for innovation. Smaller-sized models, MOE, and agents are all evolving very quickly. We strive to make these offerings more accessible to all types of enterprises and solve real-world problems in various scenarios. In the realm of visual foundation models, notably, a significant application with vast market potential is autonomous driving, in which Baidu is a pioneer and global leader. We have been using diffusion and transformers to train our video generation models for self-driving purposes. We have made strides in object classification, detection, and segmentation, which better our understanding of the physical world and the rules of the physical world. This has enabled us to translate images and videos captured on the road into specific tasks, resulting in more intelligent, adaptable, and safe autonomous driving technology. Overall, our strategy is to leverage the most powerful foundation models to solve real-world problems, and we continue to invest in this area to ensure our leadership position. Thank you.

Operator, Operator

Thank you. And ladies and gentlemen, at this time, we will conclude the question-and-answer session and conclude Baidu's fourth quarter and fiscal year 2023 earnings conference call. We do thank you for attending today's presentation. You may now disconnect your lines.