Agora Inc (API) 2025 Q3 法說會逐字稿

Good day, and thank you for standing by. Welcome to Agora, Inc., third quarter 2025 financial results conference call. (Operator Instructions) Please be advised that today's conference is being recorded.

您好，感謝您的耐心等待。歡迎參加 Agora 公司 2025 年第三季財務業績電話會議。（操作人員指示）請注意，今天的會議正在錄音。

The company's earnings results press release, earnings presentation, SEC filings and a replay of today's call can be found on its IR website at investor.agora.io.

該公司的盈利結果新聞稿、盈利演示文稿、提交給美國證券交易委員會的文件以及今天電話會議的錄音都可以在其投資者關係網站 investor.agora.io 上找到。

Joining me today are Tony Zhao, Founder, Chairman and CEO, and Jingbo Wang, the company's CFO.

今天與我一同出席的有公司創辦人、董事長兼執行長趙東尼，以及公司財務長王景波。

During this time, the company will make forward-looking statements about its future financial performance and future financial and future events and trends. These statements are early predictions that are based on what the company's beliefs today, and the actual results may differ materially. These forward-looking statements are subject to risks, uncertainties, assumptions and other factors that could affect the company's financial results and the performance of its business, in which the company is discussed in detail in its filings with the SEC, including today's earnings press release and risk factors and other information contained in the final prospects relating to its initial public offering. Angora, Inc., remains no obligation to update any forward-looking statements the company may make on today's call.

在此期間，本公司將對其未來的財務表現、未來財務事件和趨勢做出前瞻性聲明。這些聲明是基於公司目前信念的早期預測，實際結果可能與預測有重大差異。這些前瞻性聲明受到風險、不確定性、假設和其他因素的影響，這些因素可能會影響公司的財務表現和業務表現。該公司已在提交給美國證券交易委員會的文件中詳細討論了這些因素，包括今天的盈利新聞稿以及與首次公開募股相關的最終前景中所包含的風險因素和其他資訊。Angora, Inc. 不承擔更新公司在今天電話會議上可能作出的任何前瞻性聲明的義務。

With that, let me turn the call over to Tony. Please go ahead.

那麼，我把電話交給東尼吧。請繼續。

Thanks, operator, and welcome, everyone, to our earnings call. I'll first review our operating results from the past quarter.

謝謝接線員，歡迎各位參加我們的財報電話會議。我先回顧一下上個季度的經營業績。

We are pleased to report our fourth-consecutive quarter of GAAP profitability in Q3 supported by double-digit revenue growth and expanding margins. Total revenue in Q3 reached $35.4 million, up 12% year over year. Our GAAP net profit for the quarter was $2.7 million, with a GAAP net margin of 7.8%, and we expect our revenue and net profit to continue growing on a quarter-over-quarter basis in Q4.

我們很高興地宣布，在兩位數營收成長和利潤率擴大的支持下，我們第三季實現了連續第四個季度的GAAP獲利。第三季總營收達到 3,540 萬美元，年增 12%。本季我們的 GAAP 淨利潤為 270 萬美元，GAAP 淨利潤率為 7.8%，我們預計第四季的營收和淨利潤將繼續環比成長。

As you can see, our core real-time engagement task business is rebounding strongly and is on track to deliver its first full year revenue growth since the pandemic, providing a stable, profitable foundation for us. At the same time, we are significantly increasing our investment in conversational AI. Voice-based human machine interaction is not new, yet most conversational AI solution today still disappoint users. Why? Because building voice agent that can converge naturally with a human is just hard.

如您所見，我們的核心即時互動任務業務正在強勁反彈，並有望實現疫情以來的首個全年收入增長，為我們提供了一個穩定、盈利的基礎。同時，我們正在大幅增加對對話式人工智慧的投資。基於語音的人機互動並非新鮮事物，但如今大多數對話式人工智慧解決方案仍然無法讓用戶滿意。為什麼？因為建構一個能夠自然地與人類語音融合的語音代理實在太難了。

Just a few months ago, GreenLake partners, a leading rental capital firm, published a block post titled voice agents, easy to use, hard to build. They learned that the core challenge behind the simplicity user expect lives immerse complexity, system obstruction, real-time audio processing, latency management, and compliance requirements, consider the issue of Biganos article speakers just two of the many technical challenges in robo setting like a busy home office or car, clean audio is an exception, not the norm. A voice agent must accurately isolate the user's voice from our lapping speech and ambient funds.

就在幾個月前，領先的租賃資本公司 GreenLake Partners 發表了一篇題為「語音代理，易於使用，但難以建立」的大型文章。他們了解到，用戶期望的簡單性背後，核心挑戰在於生活所面臨的複雜性、系統障礙、即時音訊處理、延遲管理和合規性要求。以 Biganos 的文章為例，揚聲器問題只是機器人環境（如繁忙的家庭辦公室或汽車）中眾多技術挑戰中的兩個，在這樣的環境中，清晰的音頻是一種例外，而不是常態。語音代理必須準確地將使用者的聲音從周圍的聲音和環境音中分離出來。

Without this, transcription becomes unreliable, intent is misunderstood and the agent's reasoning falters undermining the whole interaction. Furthermore, as entry capacity has pointed out, there is often a significant gap between a working demo and the production-ready product.

如果沒有這一點，轉錄就會變得不可靠，意圖會被誤解，代理人的推理也會站不住腳，從而破壞整個互動。此外，正如入門級產品所指出的那樣，工作演示與量產產品之間往往存在著很大的差距。

Conversational AI is no exception. For instance, in our discussion with customers and prospects, many have expressed frustration with reliability and scalability of current solutions, especially when users are distributed across geographics or when concurrent usage is high. Our investment in conversational AI is specifically aimed at addressing these challenges.

對話式人工智慧也不例外。例如，在與客戶和潛在客戶的討論中，許多人對當前解決方案的可靠性和可擴展性表示不滿，尤其是在用戶分佈於不同地理位置或併發使用量較高的情況下。我們對對話式人工智慧的投資正是為了回應這些挑戰。

Recently, we launched our conversational AI engine 2.0. It integrates over a decade of Levante audio research and development, including AI-powered noise suppression, acoustic eco-concelation proprietary audio codec and adaptation across thousands of device types to ensure that AI hears and speaks with consistent quality. In addition, the engine also tackles core interaction, talent, selective attention, per-packing interruption handling, emotion detection, and natural conversational flow. In short, we're not just providing the transmission pipeline for voice and video, we're building the behavioral intelligence that powers truly responsive human-like conversational AI agents.

最近，我們推出了對話式 AI 引擎 2.0。它整合了 Levante 十多年的音頻研發成果，包括 AI 驅動的噪音抑制、聲學化學補償專有音頻編解碼器以及跨數千種設備類型的適應性，以確保 AI 能夠以一致的品質聽和說。此外，該引擎還處理核心互動、人才、選擇性注意力、每次打包中斷處理、情緒檢測和自然對話流程。簡而言之，我們不僅為語音和視訊提供傳輸管道，還建立了驅動真正響應迅速、類人對話式人工智慧代理的行為智慧。

To help developers build worse agents more easily, we announced our conversational AI studio at our recent convo AI and RTE contracts in last October -- in late October, which allows developers to create, configure and deploy voice agent through a z-code interface, Complementing this, our concessional AI benchmark and our station platform allows developers to evaluate, mix and match and optimize both our proprietary and third-party modules, so they can identify the best-performing combination for their specific use case.

為了幫助開發者更輕鬆地建立更差的代理，我們在去年十月下旬的對話式人工智慧和即時增強技術合約會議上宣布了我們的對話式人工智慧工作室。該工作室允許開發者透過 z-code 介面建立、配置和部署語音代理。此外，我們的優惠人工智慧基準測試和工作站平台允許開發者評估、混合搭配和優化我們的專有模組和第三方模組，以便他們能夠找到最適合其特定用例的最佳組合。

Our insource time framework designed for building with agents continues to gain traction in the developer community, recognized for its high concurrency architecture and deep cross-platform integration, it has been adopted by multiple cloud providers and major enterprises for their agent attrition platforms. All these products are backed by our global distributed real-time insurance cloud over the past several months, we expanded this infrastructure to cover key regions across North America, South America, Europe, and Asia, ensuring consistent latency, reliability and performance even under high concurrency and worrying network conditions.

我們專為與代理商一起建置而設計的內部時間框架在開發者社群中持續獲得認可，其高並發架構和深度跨平台整合得到了廣泛認可，已被多家雲端供應商和大型企業採用到其代理流失平台中。所有這些產品都由我們全球分散式即時保險雲端提供支援。在過去的幾個月裡，我們擴展了這項基礎設施，涵蓋了北美、南美、歐洲和亞洲的關鍵地區，即使在高並發和令人擔憂的網路條件下，也能確保一致的延遲、可靠性和效能。

Early adoptions from customers around the world has been encouraging, and our pipeline of use cases and prospects continue to grow as we head into next quarter. Our recent color AI and RTE contracts attracted more than 3,000 on-site attendees, a record for us and made it the largest gathering focused on conversational AI technology globally. Our customer and developers are deploying our conversational AI solutions to build voice agent for outbound marketing, inbound customer service, tutoring and many other applications.

來自世界各地客戶的早期採用情況令人鼓舞，隨著我們進入下一季度，我們的用例和潛在客戶數量將繼續增長。我們最近簽署的彩色人工智慧和即時增強技術合約吸引了超過 3000 名現場與會者，創下了我們的紀錄，使其成為全球規模最大的專注於對話式人工智慧技術的聚會。我們的客戶和開發人員正在部署我們的對話式人工智慧解決方案，以建立用於對外行銷、對內客戶服務、輔導和許多其他應用的語音代理。

However, manufacturers are also integrating our technology into smart plays, enabling voice powered companionship and learning experiences. In concluding, the convergence of advanced AI models and robust real-time infrastructure is unlocking a new era of possibilities backed by proven scalability. This technology expertise and the forward-looking product suite, we are well positioned to empower this next chapter, enabling truly human-like, reliable and scalable voice agents.

然而，製造商們也在將我們的技術整合到智慧遊戲中，從而實現語音驅動的陪伴和學習體驗。總之，先進的人工智慧模型與強大的即時基礎設施的融合，正在開啟一個充滿無限可能的新時代，並擁有經實踐驗證的可擴展性。憑藉技術專長和前瞻性的產品套件，我們已做好充分準備，開啟下一個篇章，打造真正像人類一樣可靠且可擴展的語音代理。

With that, let me turn things over to Jingbo, who will review our financial results.

接下來，我將把發言權交給景波，他將審查我們的財務表現。

Thank you, Tony. Hello, everyone. Let me start by first reviewing financial results for the third quarter of 2025. And then I will discuss the outlook for the fourth quarter. Total revenues for the third quarter reached $35.4 million, up 12% year over year, representing a third consecutive quarter of double-digit organic growth.

謝謝你，托尼。大家好。首先，讓我回顧一下2025年第三季的財務表現。接下來，我將討論第四季的展望。第三季總營收達到 3,540 萬美元，年增 12%，連續第三個季度實現兩位數的有機成長。

If we look at the two business divisions, Agora revenues reached $18.2 million in Q3, representing 15.9% year-over-year growth and flat quarter-over-quarter. The strong year-over-year growth reflects our successful market penetration and growing adoption in verticals such as live shopping. Shengwang revenues reached RMB122.4 million in Q3, up 8.4% year over year and 6% sequentially, driven by continued business expansion, and adoption in key verticals such as social, entertainment and IoT. Dollar-based net retention rate is 108% for Agora and 90% for Shengwang, marking the fourth-consecutive quarter of improvement for both businesses.

如果我們看一下這兩個業務部門，Agora 第三季的營收達到 1,820 萬美元，年增 15.9%，環比持平。強勁的同比增長反映了我們成功的市場滲透以及在直播購物等垂直領域日益增長的採用率。盛望科技第三季營收達人民幣1.224億元，年增8.4%，季增6%，主要得益於業務持續擴張以及在社交、娛樂和物聯網等關鍵垂直領域的應用。以美元計，Agora 的淨留存率為 108%，Shengwang 的淨留存率為 90%，兩家公司連續第四個季度實現成長。

Gross margin for the third quarter was 66%, slightly decrease 0.7% year over year and 0.8% sequentially. Moving on to expenses. R&D expenses were $13.8 million in Q3, decreased 52.8% year over year. R&D expenses represented 39.1% of total revenues in the quarter compared to 92.7% in Q3 last year.

第三季毛利率為 66%，年減 0.7%，季減 0.8%。接下來是費用部分。第三季研發費用為1,380萬美元，年減52.8%。本季研發支出佔總營收的 39.1%，而去年第三季則為 92.7%。

Sales and marketing expenses were $6.5 million in Q3, decreased 5.6% year over year. So the marketing expenses represented 18.3% of total revenues in the quarter compared to 21.7% in Q3 last year. G&A expenses were $5 million in Q3, decreased 48.4% year over year. G&A expenses represented 14.1% of total revenue in the quarter. compared to 30.8% in Q3 last year.

第三季銷售和行銷費用為 650 萬美元，年減 5.6%。因此，本季行銷費用佔總收入的 18.3%，而去年第三季為 21.7%。第三季一般及行政費用為 500 萬美元，年減 48.4%。本季一般及行政費用佔總營收的14.1%，而去年第三季為30.8%。

Moving on to the bottom line. We delivered net income of $2.7 million in Q3, representing a 7.8% net income margin. This result represents a significant improvement from last year and marked our fourth consecutive quarter of GAAP profitability.

接下來是結論部分。第三季度，我們實現了270萬美元的淨收入，淨利潤率為7.8%。這項結果較去年有了顯著改善，也標誌著我們連續第四個季度實現GAAP獲利。

Based on our current business momentum and visibility into the fourth quarter, we expect net income to grow sequentially compared to Q3. Now turning to cash flow. Operating cash flow was $0.7 million in Q3 compared to negative $4.6 million last year.

根據我們目前的業務發展勢頭和對第四季度的可見性，我們預計淨利潤將比第三季度環比增長。現在來看現金流。第三季經營現金流為 70 萬美元，去年同期為負 460 萬美元。

Moving on to balance sheet. We ended Q3 with $374.3 million in cash, cash equivalents, bank deposits and financial products issued buybacks. Net cash outflow in the quarter was mainly due to share repurchase of $4.8 million. In the third quarter, we repurchased 5.2 million ordinary shares or 1.3 million ADSs, representing 1.4% of our outstanding shares at the beginning of the quarter.

接下來來看資產負債表。第三季末，我們持有現金、現金等價物、銀行存款和已發行回購的金融產品共 3.743 億美元。本季淨現金流出主要是由於回購股票 480 萬美元。第三季度，我們回購了 520 萬股普通股或 130 萬股美國存託股份，佔季度初我們已發行股份的 1.4%。

Since our Board approved a share repurchase program in February 2022, we have repurchased $132.1 million worth of shares through September 30, 2025. The share repurchase program demonstrates our dedication to returning value to our shareholders, balanced with our ability to continue investing in strategic growth opportunities.

自董事會於 2022 年 2 月批准股票回購計畫以來，截至 2025 年 9 月 30 日，我們已回購價值 1.321 億美元的股票。股票回購計畫反映了我們致力於為股東創造價值的決心，同時也反映了我們持續投資策略成長機會的能力。

Now, turning to guidance for the fourth quarter of 2025. We currently expect total revenues to be between $37 million and $38 million compared to $34.5 million in the fourth quarter last year, representing year-over-year growth rate of 7.2% to 10.1%. This all reflects our current and preliminary views on the market and operational conditions, which are subject to change.

現在，讓我們來看看2025年第四季的業績展望。我們目前預計總收入將在 3,700 萬美元至 3,800 萬美元之間，而去年第四季為 3,450 萬美元，年成長率為 7.2% 至 10.1%。以上內容反映了我們目前對市場和營運狀況的初步看法，但這些看法可能會改變。

In closing, I would express my gratitude to our outstanding team in core and chair, sustained double-digit revenue growth and profit expansion direct reflection of your hard work and strategic focus. To our shareholders, thank you for your continued trust. We remain focused on executing on road map to build a durable market-leading company and forefront of AI innovation.

最後，我要對我們核心團隊和主席團隊的傑出貢獻表示感謝，持續兩位數的收入成長和利潤擴張直接體現了你們的辛勤工作和策略重點。感謝各位股東一直以來的信任。我們將繼續專注於執行路線圖，打造一家持久的市場領先公司和人工智慧創新前沿企業。

Thank you all for joining today's call. Let's open it up for questions.

感謝各位參加今天的電話會議。現在開始接受提問。

(Operator Instructions) Harry Zhuang, BofA Securities.

（操作員說明）Harry Zhuang，美國銀行證券。

Congratulations on another quarter of double-digit growth and a solid guidance for the fourth quarter this year. I have three questions. First is regarding the demand outlook, comanage share about the key trends in both domestic and international markets for the coming quarters? And what are the key downstream sectors that are driving the demand growth?

恭喜公司連續第二季實現兩位數成長，並對今年第四季業績給出了穩健的預期。我有三個問題。首先是關於需求前景，請問各位能否分享未來幾季國內外市場的主要趨勢？那麼，哪些關鍵的下游產業正在推動需求成長？

Second question is regarding the AI application. Could management share the latest update on the development of AI and also what are the key scenarios that could drive meaningful revenue contribution in the near term?

第二個問題是關於人工智慧應用的。管理階層能否分享人工智慧發展的最新進展，以及哪些關鍵場景可能在短期內帶來顯著的收入貢獻？

And thirdly is on the profitability outlook. Could management share the profitability outlook for both fourth quarter this year and also for '26, the operating profit level and also the net profit level.

第三點是關於獲利前景。管理階層能否分享今年第四季以及2026年全年的獲利前景，包括營業利潤水準和淨利水準？

All right. I'll take the first question and Jingbo will take the last one. So for the demand in China, the overall demand recovery trend continues. With stabilized regulatory environment, and demand from social entertainment and education customers rebounded and gradually goes up.

好的。我來回答第一個問題，景波回答最後一個問題。因此，就中國市場而言，整體需求復甦趨勢仍在持續。隨著監管環境趨於穩定，社交娛樂和教育類客戶的需求反彈並逐步上升。

Demand from IoT and digital transformation customers are experiencing rapid growth. In US and international markets, live commerce demand continued its rapid growth, and other verticals generally show growth as well. The overall growth rate is slightly faster than in China.

來自物聯網和數位轉型客戶的需求正在快速成長。在美國和國際市場，即時電商需求持續快速成長，其他垂直領域也普遍呈現成長態勢。整體成長速度略高於中國。

As to the AI demand and the trend, before I answer the question, I want to first clarify the difference between voice AI and the conversational AI. We are actually focused on conversational AI, which is very related to our real-time engagement business. And it means the real-time human AI voice interaction.

關於人工智慧的需求和趨勢，在回答這個問題之前，我想先澄清一下語音人工智慧和對話式人工智慧之間的差異。我們實際上專注於對話式人工智慧，這與我們的即時互動業務密切相關。這意味著即時的人機語音互動。

On the other hand, voice AI is a much broader concept. It includes both real-time conversation and no real-time functionalities such as audio recognition and generation. The real-time use cases actually much more broad. And that real-time audio recognition generation is also much more easier to achieve usability and to find practical use cases.

另一方面，語音人工智慧是一個更廣泛的概念。它既包含即時對話功能，也包含非即時功能，例如音訊辨識和生成。即時應用場景實際上要廣泛得多。而且，即時音訊辨識產生也更容易實現可用性，並找到實際應用案例。

So in the past two years, audio generation or tax speech has been widely used in real-time content production. For example, most of the short video clips people watch today are using AI-generated voice over. So those have been growing in the last two years in the social media and a lot of other markets. However, when we move to real-time conversation, the complexity of the technology makes the whole experience much more challenging as I stated in the opening remarks and takes longer to mature and gain adoption.

因此，在過去的兩年裡，音訊生成或語音合成技術已廣泛應用於即時內容製作。例如，如今人們觀看的大多數短影片都使用了人工智慧生成的配音。因此，在過去兩年裡，這些領域在社群媒體和許多其他市場都取得了成長。然而，當我們轉向即時對話時，技術的複雜性使得整個體驗更具挑戰性，正如我在開場白中所述，並且需要更長的時間才能成熟和被接受。

In conversational AI applications, currently, there are three use cases have progressed to more advanced stage, namely call center, education and companionship toys. For these use cases, we already see some customers have moved from proof of concept phase to real world production.

在對話式人工智慧應用中，目前有三個用例已經發展到更進階的階段，分別是呼叫中心、教育和陪伴玩具。對於這些應用案例，我們已經看到一些客戶已經從概念驗證階段過渡到實際生產階段。

Given the vast scale and the potential usage of these verticals, we expect the success of these customers will drive broader adoption. We already have customers in production today, but due this is still ramping up, we expect to see some sizable conversational AI revenue in the first half of next year, and hopefully, conversational AI will become a meaningful revenue contributor towards the end of next year.

鑑於這些垂直行業規模龐大且具有巨大的潛在用途，我們預計這些客戶的成功將推動更廣泛的採用。我們目前已經有客戶在生產中使用我們的產品，但由於這項技術仍在發展中，我們預計明年上半年對話式人工智慧的收入將相當可觀，並希望到明年年底，對話式人工智慧能夠成為重要的收入來源。

Okay. Third question. So for Q4 this year, given that Q4 is normally a strong season for us. We expect to achieve GAAP operating profit per case in Q4. And therefore, the GAAP net project will further grow on top of the Q3 level.

好的。第三個問題。所以對於今年的第四季來說，考慮到第四季通常是我們業績強勁的季度。我們預計第四季將實現每箱GAAP營業利潤。因此，GAAP淨項目將在第三季的基礎上進一步成長。

So for next year, our target is to achieve GAAP operating profit for the full year of 2026. GAAP net profit first of all, a big improvement over 2025 already. In terms of the GAAP net profit, that will have some level of uncertainty due to the potential interest rate cut.

因此，我們明年的目標是實現 2026 年全年 GAAP 營業利潤。首先是GAAP淨利潤，比2025年已經有了很大的改善。就 GAAP 淨利潤而言，由於潛在的降息，這將存在一定程度的不確定性。

But under the current forecast, we expect year-over-year net income improvement over 2025 as well.

但根據目前的預測，我們預計到 2025 年淨利也將逐年成長。

[Rachel Han], CICC.

[Rachel Han]，CICC。

Can you hear me?

你聽得到我嗎？

Yes.

是的。

This is [Rui] Han from CICC. First of all, congrats on the solid growth this year and especially the continued improvement in profitability. I have two questions. First, I noticed that our third-quarter revenue came in slightly above the midpoint of the guidance range. Could you give us more color on what drove the solid performance?

這是中金公司的韓瑞。首先，恭喜公司今年穩健成長，尤其恭喜獲利能力持續提升。我有兩個問題。首先，我注意到我們第三季的營收略高於預期範圍的中點。能否詳細介紹促成這次出色表現的因素？

And my second question is on the AI side, which downstream applications are showing the strongest momentum so far. In particular, how is the adoption train for AI companionship toys and when should we expect these use cases to start contributing to your financial results.

我的第二個問題是關於人工智慧方面的，目前哪些下游應用的發展勢頭最強？具體來說，人工智慧陪伴玩具的普及情況如何？我們預期這些應用案例何時才能開始對貴公司的財務表現做出貢獻？

Okay. I'll take the first question. So as Tony mentioned in earlier question. So for Q3, actually, we see pretty strong demand from US and international market as well as the China market.

好的。我來回答第一個問題。正如托尼在之前的問題中提到的那樣。所以實際上，在第三季度，我們看到來自美國、國際市場以及中國市場的需求都非常強勁。

In the US international market, live commerce continues to grow very strongly, and especially in more developed markets. And the other verticals such as social and fintech also growing pretty well in Q3. And in China, first of all, we had a summer vacation in 2003, which is generally a strong season for social apps and also full education apps, and in addition, the IoT sector. So smart cameras, smart wearable devices like watches, and also the smart hoist experiencing very rapid growth.

在美國國際市場上，活體電商持續強勁成長，尤其是在較發達的市場。其他垂直領域，如社交和金融科技，在第三季也成長得相當不錯。在中國，首先，2003 年我們迎來了暑假，這通常是社交應用、教育應用以及物聯網產業的旺季。因此，智慧相機、智慧型穿戴裝置（如智慧手錶）以及智慧起重機都經歷了非常快速的成長。

Yes. So for the AI sort of use cases side, they're quite strong pipeline of customers and prospects for call centers, including outbound marketing and inbound customer services. For AI companionship toys, we see strong momentum from our customer, [Robo-Pong]. Their sales and usage number are quite impressive, and they also started to charge end user monthly subscription fees, which we believe is a more healthy and sustainable business model and also a breakthrough in similar kind of toys.

是的。因此，就人工智慧應用案例而言，呼叫中心擁有相當強大的客戶和潛在客戶管道，包括外呼行銷和入站客戶服務。對於人工智慧陪伴玩具，我們看到了來自客戶的強勁需求。[機器人乒乓]他們的銷售和使用量相當驚人，而且他們也開始向終端用戶收取月度訂閱費，我們認為這是一種更健康、更永續的商業模式，也是同類玩具的突破。

A couple of other manufacturers are also in the process of integrating our solutions and we expect to see the coming to market to see them coming to market in the next few months.

其他幾家製造商也正在整合我們的解決方案，我們預計在未來幾個月內將產品推向市場。

[Ching], China Securities.

[Ching]，中國證券。

And congrats on the execution quarters. And so my first question relates on AR usage. Could you share the sequential growth trend for AI-related usage. And looking at our current customer pipeline, when could we see a sign of a meaningful scale for these AI applications?

恭喜你們本季執行順利。所以我的第一個問題與AR的使用有關。能否分享一下人工智慧相關應用領域的連續成長趨勢？從我們目前的客戶管道來看，我們什麼時候才能看到這些人工智慧應用達到有意義的規模？

And my second question relates on maybe strategy expansion. We could see some CDN vendors expanding into GPU in virens and security, given our RD infrastructure, can we see or a similar path to maybe offer or just cross-sell edge side ingress or security features?

我的第二個問題可能與策略擴張有關。鑑於我們的 RD 基礎設施，我們可能會看到一些 CDN 供應商將業務擴展到 GPU 領域，用於病毒傳播和安全防護。我們是否可以看到類似的路徑，例如提供或交叉銷售邊緣側入口或安全功能？

Yeah. So conversational AI usage increased by more than 150% on a quarter-over-quarter basis. So it's quite fast. And although, as I mentioned, the voice AI is mature for years already, but conversational AI is still at the early stage. And we do see a strong pipeline of customers and prospects.

是的。因此，對話式人工智慧的使用量較上季成長超過 150%。所以速度相當快。雖然正如我之前提到的，語音人工智慧已經發展成熟多年，但對話式人工智慧仍處於早期階段。我們確實看到了大量現有客戶和潛在客戶。

We believe we are not far from more broader adoption and [preliteration] of voice agents. For the second question, we're not a sealing company, but we do have a global distributed network and a large number of centers distributed across every major region. So it's a good question. In fact, we have opportunities that are similar, but from a different perspective.

我們相信，語音代理的更廣泛應用和普及（預文字化）指日可待。對於第二個問題，我們不是一家密封公司，但我們確實擁有全球分散式網絡，並在各個主要地區設有大量中心。所以，這確實是個好問題。事實上，我們面臨類似的機遇，但視角不同。

Specifically, we're targeting real-time insurance services for conversational AI. This is what we build for our own product. This inference service needs to connect with multiple distributed SR TTS large language model service as well as our self-development and deployed modules in different locations.

具體來說，我們的目標是將對話式人工智慧應用於即時保險服務。這是我們為自己的產品打造的。此推理服務需要與多個分散式 SR TTS 大型語言模型服務以及我們自行開發和部署在不同位置的模組進行連接。

This kind of capability is a mask to support the cloud process in a way that it has to be very latency so that the real time nature of the interaction could be enabled. Such infrastructure service of great value to any agent that requires ultralow latency or real-time inference. So this is also an opportunity we could expand in the future.

這種能力是一種掩飾，它以一種必須具有極低延遲的方式來支援雲端進程，從而實現互動的即時性。對於任何需要超低延遲或即時推理的代理商來說，這種基礎設施服務都具有極大的價值。所以這也是我們未來可以拓展的機會。

(Operator Instructions) There are no further questions. That concludes today's Q&A session. Thank you, everybody, for attending the company's call today.

（操作說明）沒有其他問題了。今天的問答環節到此結束。感謝各位今天參加公司電話會議。

As a reminder, the recording and the earnings release will be available on the company's website at investor.agora.io. And if there's any other questions, please feel free to email the company. Thank you.

再次提醒，會議錄音和財報將在公司網站 investor.agora.io 上發布。如有任何其他疑問，請隨時發送電子郵件至本公司。謝謝。