Vobile Group (3738.HK)
First Quarter 2026 Earnings Call
Disclaimer: This English transcript was generated via AI translation from the original Chinese investor call.
While we strive for accuracy in our communications, AI translations may occasionally contain errors or
variations in business context. The original Chinese transcript remains the definitive and authoritative
record of the meeting.
May 19th, 2026 17:00 – 18:00 HKT
Note: All amounts in Hong Kong Dollars (HKD) have been converted to United States Dollars (USD) based on the exchange
rate in effect on May 19, 2026, at 1 HKD = 0.1277 USD.
Katherine Tian, Director, Investor Relations
The company's revenue for the first quarter of 2026 achieved a year-over-year growth of 21%. The Q1
data disclosed in the announcement represents unaudited operating revenue, amounting to
approximately USD 111.1 million. When comparing Q1 with Q4 of last year, we achieved very strong
sequential growth, reaching 44%.
AI has brought monumental changes to the content industry, making copyright issues and content
monetization needs increasingly prominent. As a global leader in copyright management, our
performance this quarter reflects the company's robust business growth and strong market demand.
Breaking down the revenue by region, the disclosed revenue growth rate for the China region is 22%, with
a revenue scale of approximately USD 52.4 million.
Although the revenue growth rate and scale for the US region were not explicitly disclosed, a simple
calculation shows that the US region's revenue growth is nearly 20%. Its scale exceeds that of the China
region, reaching approximately USD 58.8 million.
Due to adjustments in AI content monetization rules in Q4 of last year, revenue from AI-monetized assets
slowed during that period. However, this year, with the massive adoption of AI and the substantial output
of content across the industry, the company has achieved breakthrough growth in AI asset monetization.
Based on the active asset management data within our value-added revenue, the company's active assets
reached 5.24 million in Q1 2026, representing a 32% year-over-year increase. Compared to the recently
disclosed 4.86 million active assets in Q4 2025, Q1 saw a sequential increase of 387,000 active assets.
Notably, AI-related active assets accounted for 16.6% of the total, reaching 870,000.
Furthermore, our Monthly Recurring Revenue (MRR) grew by 29%. This indicates that a significant portion
of our business growth continues to be driven by stable, sustainable revenue streams.
Jack Long, Senior Vice President
I will share our latest progress regarding business development from two main aspects: first, changes in
our revenue composition, and second, advancements in AI-related revenue.
Briefly reviewing last year's situation: in Q4 2025, the establishment of certain rules between content
owners and large language model (LLM) providers was still ongoing. Consequently, our value-added
business on YouTube was temporarily impacted by these rules, as some content was used as evidence in
infringement litigation rather than being monetized. This was the backdrop for the fourth quarter of last
year.
During our 2025 annual results call, we shared a market observation: we believed that within about a
year, a consensus on AI content rules would be reached among major copyright owners, model providers,
and platforms. Recently, we witnessed a significant development: on April 21, YouTube officially
announced its support for identifying AI deepfake content. Specifically, for AI-generated content featuring
famous or recognizable portraits/images, creators or portrait owners can now submit takedown requests
to YouTube. Essentially, YouTube has initiated a large-scale cleanup of fake celebrity or prominent IP
videos generated by AI.
This tool is called Likeness Detection. Its underlying logic is highly similar to YouTube's previous Content ID
copyright recognition system. When YouTube launched Content ID around 2013, the initial process also
allowed copyright owners to identify infringing content and issue takedowns; the subsequent step was
that if the copyright owner opted not to take the content down, they could share in the advertising
revenue. The current approach is strikingly similar. Platforms like YouTube have fundamentally
acknowledged that deepfakes or copyright abuses generated by AI constitute infringement. We believe
that, from a commercial standpoint, this will soon transition into a monetization phase.
While anticipating this shift, the company took two strategic actions at the business level:
First, regarding social media operations, primarily on YouTube: if certain infringing content we previously
targeted is now tied up in bilateral litigation, our response has been to expand our monitoring scope. This
action has driven supplementary growth in the active assets we manage on YouTube.
Second, throughout last year, the AI-related assets we managed were predominantly AI-driven secondary
creations, such as end-frame continuations and reused copyrighted clips. In the first quarter of this year,
our business team significantly intensified efforts to acquire AI-original content, such as AI-generated
comic videos and live-action shorts, and distributed them across social media platforms. In terms of
production cost per unit of time, AI-native content is substantially cheaper than traditional filming;
however, when monetized on social media, its earning potential per unit of time is virtually identical.
Consequently, this content segment has unlocked massive potential for the company.
Therefore, the nearly 17% contribution from AI original content to our active assets mentioned earlier is
largely driven by this AI-native content. The company is actively benefiting from the rapid development of
multimodal technologies. By broadening our overall monitoring and rights-confirmation scope, and
capitalizing on the growth of AI-native content, you can see that we delivered highly robust year-over-year
revenue growth in Q1.
In addition to proactively acquiring AI-generated content, the DreamMaker ecosystem we have been
developing has also attracted—or rather, we have selectively invited—a substantial number of
professional creators to the platform. The short films, music, MVs, narrative content, and mini-dramas
produced on these platforms have all successfully entered our monetization framework.
In Q1 2026, by inviting creators to utilize the DreamMaker platform's services, the company generated
revenue of nearly HKD 30 million (approximately USD 4 million), which is roughly equivalent to our fullyear revenue for this segment in 2025. We initially launched this platform in 2025, during which time it
was in a trial-and-optimization phase with targeted invite-only usage.
By successfully integrating our proprietary copyright services with AI models, we are providing
professional computing power services and clearing the path for copyright monetization—a strong
demand we are seeing across the industry. Moving forward, the company will roll out further initiatives
aimed at building a deeply integrated ecosystem combining computing power, user-friendly commercial
models, and copyright services. This includes our strategic layout for overseas computing power
partnerships, which is currently a key focus area.
Regarding our China operations, as mentioned in our previous annual results call, the company is steadily
deepening its copyright service partnerships with ByteDance products, such as Hongguo and Qishui
Music, including collaborations on more innovative product fronts. In China's AI content production
landscape, we are also observing the emergence of "AI content factories" and local industrial clusters.
Some of these entities are already producing tens of thousands of minutes of content per month, or even
more. Addressing the copyright protection and monetization needs of these companies—by combining
computing power, high-efficiency models, and, crucially, our globally distributed copyright service
system—represents a massive opportunity for the company in China.
Q&A Session
Question 1:
The company mentioned the revenue performance of the DreamMaker
platform. When do we expect to see a significant volume ramp-up for this
business? Does the platform compete with other existing multimodal content
platforms, and what are its competitive advantages?
Jack Long：
I'll take this opportunity to provide some additional context on DreamMaker.
DreamMaker essentially aggregates a wide array of multimodal models, spanning
both video and audio dimensions, integrating them to attract and serve
professional creators.
Content generated through DreamMaker immediately receives copyright services
from Vobile. Once the content is created, copyright registration is executed
instantly—including the application of fingerprinting and watermarking
technologies—and we facilitate its distribution to global social media platforms
for monetization. Within this ecosystem, the massive volume of content
produced by large AI models can achieve maximized content liquidity, tracking,
and revenue aggregation. This is the core business DreamMaker is currently
executing.
Last year, the company initially piloted this business on a small scale utilizing
computing power supported by the University of Florida. Setting aside the
monetization revenue from the content itself, simply looking at the utilization of
DreamMaker's computing power—since computing power and services
inherently carry costs—the company bundled this with professional copyright
services to generate revenue with a solid profit margin.
As the platform has steadily rolled out in North America, we have successfully
connected with a growing base of professional creators, including the
partnerships with professional Hollywood studios we announced last year. In
practice, the platform is already being utilized across multiple sectors. In Q1 of
this year alone, DreamMaker generated nearly HKD 30 million (approximately
USD 4 million) in computing power revenue tied to professional content
production. Once this content enters the distribution and monetization
pipeline—meaning it is concurrently published and monetized on platforms like
YouTube and TikTok—it generates yet another stream of revenue for our valueadded services segment.
Therefore, at this level, the company has clearly validated the creators' demand
for computing power, their need for seamless workflows that integrate multiple
models for content production, and their desire for diversified content
monetization. Against this backdrop, the company is now actively deploying
large-scale computing infrastructure beyond the University of Florida, specifically
and vertically tailored to serve video and audio content production.
In other words, once this system is fully operational, computing power, models,
and our corporate services will be organically integrated. Based on our current
progress—while some details are not yet suitable for public disclosure today—we
expect our overseas computing power infrastructure to begin generating
meaningful revenue in the second half of this year. As for the overall trajectory,
you have already heard the Q1 revenue figures. The company anticipates that
the DreamMaker-related business will truly hit its inflection point in scale during
the second half of 2026, delivering substantial contributions during that phase.
Question 2:
Management previously mentioned that the RWA project was slated for
issuance in the first half of the year. With regulatory frameworks like Hong
Kong's stablecoin licensing gradually falling into place, the policy environment
appears clearer. What is the latest timeline and pace for the RWA issuance, and
what specific plans does management have?
Jack Long：
The RWA project continues to advance under Hong Kong's regulatory framework.
As of now, all critical compliance processes have been completed. A reasonable
expectation is that the project will be officially issued in the first half of this year,
specifically in June.
To provide some additional color: in terms of scale, the company's Phase I RWA is
expected to be an asset issuance in the tens of millions of US dollars. Regarding
the underlying assets, this aligns with the consensus we reached after multiple
discussions with relevant regulatory bodies in Hong Kong. Single-IP financing is
not where the true value of RWA lies. You may have noticed in relevant RWA
guidelines that there are fundamental requirements regarding whether a
sustainable revenue stream is suitable as an RWA asset.
Within the broader film and television industry, the revenue stream generated by
Vobile’s content services every day, hour, and minute is actually a continuous
cash flow aggregated from diverse forms of content. This includes long-form and
short-form content, domestic and international IP, audio, and video. Compared
to a single film or series, a revenue stream derived from diversified content and
dynamic management significantly mitigates the volatility of revenue
expectations, offering far stronger overall risk resistance.
A single movie has the potential to become a blockbuster, but there is a higher
probability that it may fail to recoup its investment. Therefore, a diversified
portfolio of content assets is actually the type of asset profile that investors
prefer.
Consequently, when the RWA is launched, you will see that the underlying assets
include not only overseas film and television works but also premium Chineselanguage content, as well as highly popular micro-dramas and AI-related content.
This portfolio structure is designed to progressively pioneer an innovative
monetization model—or a novel financing solution—for new content formats,
particularly AI-generated content.
The company looks forward to hosting a dedicated discussion and sharing more
details with everyone in June, based on the specific use cases.
Question 3:
Is the company currently observing any changing trends in gross margins and
net profit margins? Additionally, Mr. Long just highlighted the explosive growth
trends in AI-native video. Will AI also provide clear empowerment and
efficiency gains for the company's own internal operations? For instance, can AI
bring tangible efficiency improvements in video scanning, recognition, or other
operational processes?
Jack Long：
The answer is absolutely yes. AI is indeed providing significant benefits both in
terms of driving revenue and enhancing our internal operational efficiency. Let's
address profitability first. If you have been tracking the company over the past
few reporting cycles, you will have noticed a continuous upward trend in both
our gross margin and net margin levels. The company expects to sustain this
trajectory throughout this year.
Beyond unlocking massive business opportunities, AI has also demonstrably
improved our internal operating efficiency. In February of this year, we
specifically showcased our element-level management capabilities to the market.
The ongoing lawsuits and disputes between major content owners and YouTube
that we discussed earlier actually translate into a direct service that content
owners are commissioning from Vobile. A key area where budgets are increasing
is utilizing these element-level tracking capabilities to help content owners trace
how their copyrighted materials are being disseminated on platforms like
YouTube after being repurposed by users via AI creation tools.
This segment differs somewhat from our traditional business. Historically, we
relied more on general computing power to execute comparisons at the
fingerprint technology level. Element-level recognition, however, requires AI
computing power and entails utilizing more complex functions, such as video
language models and advanced video comprehension. Therefore, at this stage,
this capability does introduce a shift in our cost structure compared to our legacy
operations.
To address this shift, the company is actively pursuing deep partnerships on the
computing power front. On one hand, we need robust computing infrastructure
to satisfy the surging user demand on DreamMaker. On the other hand, if we can
leverage this same computing power for our internal operations, it will
tremendously benefit our cost control and bottom-line profitability.
AI empowers the company across multiple dimensions. As a technology-driven
enterprise, our engineering headcount has not expanded significantly over the
past year or two, yet our revenue has continued to grow—partly thanks to the
efficiency gains AI brings to our coding processes. Throughout our business
operations, from operations to monitoring, AI is dramatically empowering our
workflows via Agent-based solutions. Furthermore, in the realm of element-level
management, AI computing power has already substantially elevated our
operational efficiency.
Question 4:
Regarding the convertible bonds maturing in September this year, could
management provide an update on the progress of the debt-to-equity
conversion? If the company opts for repayment, will interest need to be paid,
and at what rate? Does the company have any other strategic plans for these
convertible bonds?
Jack Long：
There is still some time remaining before the conversion deadline, which is why a
significant portion has not yet been converted. Naturally, the conversion is also
linked to specific stock price triggers. Based on the robust fundamentals of our
overall business development, the company remains highly confident.
There are many investors on the line today who know the company very well and
have been following us for over 3 to 5 years. If you review the transcripts of our
past market communications, you will find that the evolving development
strategies we present are almost always executed and realized. Therefore, the
company is fully confident that the AI strategies we’ve communicated over the
past year or two, the computing infrastructure plans I just outlined, and
initiatives like the RWA will all see substantial materialization—or strategic
realization—in the coming months.
On the flip side, looking at our cash reserves, the company is currently in its
strongest cash position in its history. Concurrently, the company is consistently
profitable. Thus, from an overall capital and liquidity perspective, we operate
from a position of profound strength.
In the context of the broader AI era—particularly the rapid evolution of
multimodal technologies—I believe everyone can intuitively sense that video
generation is genuinely entering an industrial-scale production phase. As we saw
in Q1, the revenue generated by the computing power utilized on DreamMaker
in just one quarter matched the total for all of last year.
Taking all these perspectives into account, the company believes we are entering
a highly favorable phase of rapid business acceleration. These milestones will
also allow the market to more clearly recognize the company's true value.
Question 5:
The company's performance is on a continuous uptrend, and external policy
shifts are also serving as tailwinds for the business. Have your growth
expectations for 2026 and beyond been revised upwards due to YouTube's
policy changes? What is your current revenue growth expectation for 2026?
Jack Long：
Today, we discussed a wealth of progress regarding our business operations. We
also shared specific data points—such as the DreamMaker metrics—for
initiatives that you previously knew were strategic priorities but lacked
quantifiable visibility. From this vantage point, we believe that as our business
continues its robust trajectory, we also look forward to delivering positive
surprises. We want to demonstrate to everyone that the company is operating
on a high-growth track, particularly given the massive market space unlocked by
the AI revolution.
