Agora (API) Q1 2026 earnings call transcript

Joining me today are Tony Zhao, founder, chairman, and CEO, Jimbo Wang, the company's CFO. So, during this call, the company will make forward-looking statements about its future financial performance and other future events and trends. These statements are only predictions that are based on what the company believes today. 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, and in which the company discussed in details in its filings with the SEC, including today's call, earnings press release, and their risk factors, and other information contained in the final prospects relating to its initial public offering.

Agora, Inc. remains no obligation to update any forelooking statements the company may make on today's call. Now, with that, let me turn it over to Tony.

Hi, Tony. Thank you, operator, and welcome everyone to our earnings call. I will start with a review of our operating results for the quarter. I am pleased to report our sixth consecutive quarter of gap profitability alongside another quarter of strong top-line growth. Total revenue for the first quarter of 2026 reached $37.7 million, up 13.5% year-over-year. Growth further accelerating our prior quarters, Gap net profit was $1.1 million, more than double the level of Q1 last year. These results reflect the continued expansion of our real-time engagement use cases globally. as well as the increasing contribution from AI-related applications and products built with our solutions.

Now, let me turn to our business, product, and technology update for the quarter. Over the past several months, we continue to make progress in bringing conversational AI to real-world production, Deepening the capabilities with our real-time engagement infrastructure and expanding our ecosystem partnerships. In March, we officially launched an agent studio, a visual no-code environment that enables developers and enterprise to rapidly build, test, and deploy with AI agents at scale. We also introduced conversational AI agents for inbound use cases such as customer service, as well as outbound use cases focused on sales and marketing.

The market opportunity here is enormous. According to Garner, conversational agents are expected to automate 70% of customer interactions by 2027, and by 2028, AI agents are projected to outnumber human sellers by 10 to 1.

At the same time, many enterprises still struggle to deploy voice AI in production environments. The challenge is not simply the AI model itself, but the complexity of operating multiple technology layers while maintaining low latency, reliability, and the natural conversational experience at scale. In addition, effective enterprise deployment requires domain-specific expertise. A successful voice AI agent must do more than respond accurately and must reflect the tone, personality, and workload of the industry it serves. For example, a car sales assistant and the debt collection agent need very different conversational styles, compliance scenarios, and customer engagement approaches. Our solution is designed to eliminate this complexity through a fully integrated stack that combines three core components.

that allows enterprise to design, test, and deploy AI agents in minutes rather than weeks or months. Second, our conversational AI agent outstries ASR, large-longer model, and TES capabilities with intelligent interruption handling, noise suppression, multilingual support, and domain-aware conversation design, One, enabling more natural and human-like interactions.

Third, our global real-time network infrastructure delivers sub-second latency and carry a great reliability worldwide. We are already seeing strong early validation from real-world deployments. In Q1, one customer implemented a survey and pooling agent that matched 10% conversion rate of human agents. This allowed them to scale data collection and reward distribution far more cost-effectively, Without adding operational hack-ons. Overall, enterprise feedback has been highly encouraging. Customers increasingly recognize that scalable conversational AI requires not only powerful models, but also real-time infrastructure's capability of delivering and seamless interaction and integration. We believe we are uniquely positioned at the intersection of these capabilities. Last month, we also strengthened our position in the enterprise collaboration market with the launch of our Intelligent Meeting Engine product.

Intelligent Meeting Engine offers end-to-end encryption, flexible deployment options including on-premises and private cloud, and full data isolation to help ensure that customer meeting content remains entirely within their controlled infrastructure. At the same time, it integrates AI-powered capabilities, such as real-time transcription, translation, intelligent meeting summaries, and automated follow-up workflows that can connect with customers' existing business system. This solution addresses growing enterprise demand around content, date sovereignty, and intelligent workflow automation, and has been well received in industries including finance, government, and healthcare.

Turning to ecosystem partnerships, we continue to integrate the latest AI models such as Google's Gemini Live and XAI Squawk models into our conversational AI solutions. In particular Google has featured Agoa as a recommended partner for building real conversational AI validating our technology leadership in this space In addition we recently entered a strategic partnership with NetEase Enterprise Service Division NetEase Smart Enterprise Today, together we will provide integrated solutions spanning real-time video, content moderation, and AI agents. This partnership combines expertise in AI and content moderation with our leadership in real-time engagement infrastructure. We believe this partnership is meaningful realization of our technology from one of China's leading internet companies, while also expanding our go-to-market opportunities across education, customer service, digital entertainment, and enterprise collaboration.

Before I conclude, I want to thank the Agora and Shenlong teams for their continued dedication and execution, and thank our shareholders for their ongoing trust and support. Globally, conversational AI is rapidly moving from proof of concept to large-scale deployment. Since the official launch of our conversational AI engine product last year, usage has demonstrated remarkable momentum with over 150% sequential growth every single quarter. NFRs today are no longer asking whether they should adopt conversational AI. Instead, they are asking how to deploy it at scale with reliability, low latency, and seamless integration.

We believe our decade of experience in real-time engagement infrastructure uniquely positioned us to help customers solve exactly these challenges. With that, let me turn things over to Jimbo, who will review our financial results.

Thank you, Tony. Hello, everyone. Let me start by first reviewing financial results for the first quarter of 2026. Now I will discuss outlook for the second quarter. Starting this quarter, we have simplified a disclosure approach for revenues and active customers, And we will no longer separately disclose these metrics for Agora and Shun-Wong.

We've also refined our dollar-based net retention rate for DB and ER methodology. We now compare quarterly revenue from the same cohort of paying customers year over year to calculate DB and ER. This change aligns DB and ER more closely with our quarterly revenue growth rate, Making it easier for investors to compare the two. Total revenue for the first quarter reached $37.7 million, representing 13.5% year-over-year growth. This result exceeded the high end of our guidance range of $36 to $37 million, and reflected continued expansion and usage growth of real-time engagement services The first quarter was a 99% compared to 95% in the first quarter of 2025. Gross profit first quarter was 23.9 million, representing a 5.7% year-over-year increase.

The gross margin was 63.4% compared to 68% in the same period last year, mainly due to product mix change, especially conversational air products remaining at a subscale stage. Turning to expenses, R&D expenses were 14.4 million in Q1, up 2.9% year-over-year. R&D expenses accounted for 38.1% of total revenues, compared to 42.1% in the same period last year. The increase was primarily due to continued investment in conversational and air products. So the marketing census were 5.9 million in Q1, down 4.8% year over year. So the marketing census represented 15.6% of total revenues in the quarter, compared to 18.7% in Q1 last year. The decrease was primarily due to this plain expense management, including lower personnel and promotion census.

General and administrative expenses were $6 million in Q1, down to 0.4% year-over-year. G&A expenses represented 15.9% of total revenues compared to 18.8% in Q1 last year. The decrease was primarily due to a lower allowance for current expected credit losses, Mainly as a result of improved customer credit conditions and collection outcomes.

Moving on to the bottom line, we delivered net income of 1.1 million in QI, more than double the net income in the first quarter last year, representing a 2.9% net income margin. This marks our sixth consecutive quarter of GAP profitability and reflects continued improvement Now, operating leverage.

Now, turning to cash flow. Operating cash flow was $5.7 million in Q1, including interest received of $4.3 million, compared to $17.6 million in Q1 last year, which included interest received of $17.8 million.

Moving on to balance sheet, we ended Q1 with 366.1 million in cash, cash equivalents, bank deposits, and financial products issued by banks. That cash flow in the quarter was mainly due to share repurchase. During the quarter, we repurchased approximately 12.5 million plus the ordinary shares, or 3.1 million ATS, representing approximately 3.6% of our total outstanding shares at the beginning of the quarter for approximately $13.1 million. As of March 31st, 2006, we had repurchased 174.7 million Class A auditing ratios, or 43.7 million ADF, for approximately $156.2 million under our share repurchase program, which represented 78.1% of a $200 million share of appreciated program. The current program will expire at the end of February, 2027.

Now turning to guidance. Based on currently available information, we expect total revenues for the second quarter of 2026 to be between $39 and $40 million. Compared to 34.3 million in the second quarter of 2025, representing year-over-year growth of 13.7 to 16.6%. Notably, even at the low end of this range, we expect to deliver faster revenue growth than we did in the first quarter.

In closing, I want to thank our teams for their focused execution in the first quarter. We beat revenue guidance and nothing more than doubled year over year Our second quarter outlook also points to a further acceleration in revenue growth We will continue to invest in AI with discipline and we are confident that it will become an increasingly important driver of long growth Thank you all for joining today call Let open up for questions Thank you As a reminder to ask a question please press star 11 on your telephone and wait for your name to be announced To withdraw your question, please press star 11 again.

Thank you. On the first question, first of all, in this quarter, both the China business and the U. S. international business are growing very rapidly. So the growth rate, the US percent is still relatively faster, but the two are approaching both at very healthy rates. So in terms of the demand in both markets, I'll talk about the demand in the RTE market first, and Tony will talk about demand in the AI market. So for RT in China, demand for money, the traditional verticals, social entertainment, education, some of these verticals demanded continue to recover. And in the U. S. international markets, demand from live shopping, financial services, and gaming spaces are among the strongest.

And we have a very healthy pipeline of new customers in these verticals as well. So, well, RTE demand looks quite healthy.

About the demand on the AI side, from the beginning of this, closely watching the progress on all fronts, we were the first to introduce AI into the whole RTE technology stack and offer the first generation of products empowered by those capabilities. Since then, we've been closely working with customers on practical demand. The thing is, in the last few years, there has been a lot of hype around how AI can change people's lives, and those claims are not fake, but many of those claims are overstatements that are far ahead of what's happening on the ground. Mostly oversimplified the practical challenge and actual adoption process. Since early last year, we've been seeing demand from Call center, education, digital avatar, et cetera.

I think we talked about that last year. This is happening over the past few quarters in different regions. In each of those areas, we actually had certain partners and customers to work with them to grow into real production. With them, we made progress in the overall experience, practical token economy, and customer use case adoption. At this moment, we see fairly large demand from the call center side. As the technology of voice agent is increasingly able to communicate and resolve many communication tasks, leveraging large-language model intelligence is improving day-by-day. On IoT side, after successfully helping to launch the companion toy, Fuzuzu, similar demand is expanding. Fuzuzu's growth itself is also very promising. It can get enough monthly subscription revenue from the most sticky user group every month.

So it's not just one time sale of the hardware toy. We're seeing a similar trend in other use cases of conversational AI.

Yeah, so Tony just talked about the demand for conversational AI. So that's also part of the second question. So yes, for revenue contribution this year, I would believe call center and IoT will be the biggest contributors. And I think Tony also talked about that in his opening remarks that since we released our commercial AI engine product in March last year, its usage has been growing at more than 150% sequential growth rate every single quarter. So, also the revenue contribution at the moment is relatively low. We expect to see revenue to quickly ramp up and towards the summer around 5% revenue contribution by then. So in terms of the 2016 operating target, so given the current growth trajectory and the seasonality, we expect operating income and net income to both close, consequentially, every quarter from Q1 to Q4.

And in terms of the three-year profit, we expect the GapNet income will be significantly higher than last year. And our goal is to achieve Gap operating profit in the second half of this year.

So, first question on commerce use case. So, I want to say that in the U. S. market and probably all developed markets in general, video-based live shopping is still a very new set. So, we mentioned one last quarter. So, after that event, very successful event, actually the It's difficult in terms of user acquisition and customer uniqueness. So we continue to see growing demand from that customer. In addition, we recently won over another fast-co-end video-based e-commerce customer in the U. S. market from a competitor. And on top of that the last question was a master in the industry So now everybody in the industry is watching and several other players are trying to host similar events in the future, and we are discussing with a few of them already.

So we do expect this verticals to have a lot of room for clothes, and we are making solid progress on the front. So in terms of the business in China, as I said earlier, demand from these internet-based use cases, social entertainment, education, we see demand recovery, also stay at a moderate rate, and also from verticals such as IoT, cameras, doorbells. Cars, from wearable devices, demand from IoT is going very fast. It has been very fast in the past two, three years. And also digital transformation customers with additional AI features. We also see renewed demand growth from digital transformation, traditional enterprise customers. In terms of the I think it's certainly very helpful in its own home, but also it reflects the further consolidation of the RTE market in China.

So recently, to just give some more examples, a private competitor in this market repurchased all of its venture capital investors and started to focus more on profitability rather than scale. Right? And that is used to be a competitor, now, and we also see another cloud, large public cloud competitor, has further reduced its staff on the ERT business. So we do believe this trend of easing, this trend of consolidation will gradually help our revenue growth as well.

Sure, so first question on competition and the margin. So I will talk about RTE and Tony will talk about it. So actually I just talked a little bit about the competition in the China market. We see that the market is moving to further consolidation with more players to no longer going after scale. So, we do believe that will help with running calls as well as margin improvement in coming quarters. So, as you can see overall the cost margin in this quarter was a few compared to the same quarter last year, but that's mostly due to the initial next cost margin of the conversational air business. If the closing covers the gross margin of the core RTE businesses, they may actually be relatively steady in the first quarter.

And on the AI side, there's a lot of competition for conversational AI in Silicon Valley and the US market. The market is still at an early stage, and there are different players trying to attack it from quite diverse angles. Because it's still a growth market, so every company has a chance to attack from different angles and still making progress. We are the ones who focus more on the fundamental technology, trying to enable the most promising use cases through the ultimate quality of conversation. The customer demand is strong. As we see, the status is actually adapting to those customer demands and improving the conversational quality so that it can resolve communication tasks at a higher and better level, making it more effective.

China market is quite different. As you can see, most of the AI companies can only make a frictional of revenue in China market compared to their peers. The market is quite hard to get to at this moment. However, there is similar demand on conversational AI side, and the technology and product progress are also similar.

Okay, so in terms of the unique economics of the conversational effort, I think it is still too early to talk about For example, in the first quarter, the reason we have an active cross-margin for this project is because we have a lot of POC customers, a lot of experimentation that basically generates no revenue but has a lot of cost. So, we believe as we continue to scale as customers, move from POC to deployments and scale usage by the end of the year, we expect to see meaningful revenue contribution, and at that point, the gross margin will certainly turn positive and also will be at a healthy level. In the long run, we actually expect the conversational airbusiness to generate similar, if not higher, cost-memory than the current RT products because of one higher pricing and to the more technical sophistication and value creation for customers.

So, if the AI progress is low, like, does not need the expectation, but we have considered all the investments we need to make on that front, and it will not affect our goal of turning operating profitability in the second half of the year.