Meitu is undergoing an AI-fueled resurgence, yet significant external apprehensions linger regarding its long-term viability.

Utilizing AI-driven paid growth strategies, Meitu, which had long been in the shadows, is now stepping back into the public and investor limelight.

The upside of Meitu’s “comeback” and transformation lies in the robust revenue and paid user growth spurred by AI applications. However, on the flip side, it faces rigorous market scrutiny concerning the sustainability of its profits and long-term competitive edge.

According to the latest figures, Meitu reported a total revenue of 1.8 billion yuan for the first half of the year, marking a 12.3% year-on-year increase. Of this, revenue from its core imaging and design product segment reached 1.35 billion yuan, surging by 45.2% compared to the previous year.

The primary catalyst for this growth is AI. The integration rate of AI capabilities in the company’s offerings has soared to 90%. From “AI Flash” to “AI Outfit Change,” a series of innovative features have fueled paid subscription growth. The number of paid users has skyrocketed more than eightfold over five years, reaching approximately 15.4 million.

However, the profit metrics that investors focus on reveal a mixed picture: While adjusted net profit surged by 71.3% year-on-year to 467 million yuan, the full-year forecast from financial data provider Chaoyang Yongxu suggests that Meitu’s net profit for 2025 is expected to range between 610 million and 994 million yuan, with year-on-year changes between -24.2% and 23.5%. This indicates potential profit volatility or downward pressure in the latter half of the year.

This nuanced disparity between revenue growth and potential net profit pressure underscores the current stage of Meitu’s development: The AI subscription model is driving rapid revenue and core profit increases. However, factors such as historical financial burdens and potentially heightened marketing investments due to fierce competition are introducing uncertainty into the overall net profit outlook.

This also implies that the market is keenly monitoring its subsequent financial reports to ascertain whether the quality of growth following its “cost reduction and efficiency improvement” initiatives can continue to exceed expectations.

01

A Desperate Gambit: Lessons from a 4.2 Billion Yuan Loss

Meitu’s current “lightweight” approach in the AI sector represents a passive rebirth following a catastrophic defeat, at the cost of over 4.2 billion yuan in losses.

From 2018 to 2019, buoyed by the success of its tool applications, Meitu embarked on a frenzy of strategic diversification: The company aggressively ventured into smartphones, short videos, e-commerce, and even cryptocurrency, aiming to swiftly monetize its vast user base.

However, the smartphone business, lacking core technology and hardware expertise, became a financial drain. The short video app “Meipai” was overwhelmed by competitors like Douyin and Kuaishou. The scattered strategic focus not only failed to open new avenues but also severely diluted the company’s management resources and cash flow.

The consequences of this reckless expansion were dire. The company’s net cash outflow reached 2.3 billion yuan over two years, its stock price plummeted by over 95% from its peak, and its market value evaporated. Founder Wu Xinhong lamented during an internal review: “In those two years, we suffered enormous losses and had less than 3 billion yuan left in our accounts. If we kept messing around, Meitu would run out of resources.”

The survival crisis compelled management to undertake a “strategic liquidation.” They realized that Meitu’s core strength lay not in hardware manufacturing or platform operations but in its understanding of visual aesthetics and its ability to apply them to the masses.

Based on this realization, the company made a pivotal decision akin to “cutting off an arm to save a life”: It fully retracted its strategic focus, shut down Meipai, divested its smartphone business, liquidated its cryptocurrency holdings, and reduced its workforce from a peak of 3,300 employees to 1,300.

Simultaneously, strategic priorities were realigned around two key directions: First, to fully embrace generative AI as a potent weapon to reshape product capabilities; second, to transition the business model from the unstable advertising model to the more predictable paid subscription model.

The lessons learned from these massive losses have paved the way for Meitu’s subsequent transformation centered on AI applications.

02

90% AI Feature Penetration and 5.5% Paid Penetration

After establishing its core strategy of focusing on AI and subscriptions, Meitu swiftly implemented AI technology at the product level, with AI features becoming a series of solutions addressing user pain points.

Currently, the penetration rate of AI features in its products has reached 90%, indicating that AI is no longer a mere novelty but is fully integrated into its product lineup.

In a research report, Orient Securities highlighted that as of October 2025, Meitu XiuXiu has introduced 51 new features this year, nearing last year’s total, with AI features accounting for a staggering 96%.

This high-frequency and intensive iteration of AI features stands out compared to many established players in the image beautification and tool application sectors that emerged alongside Meitu. It reflects the greater speed and agility of Meitu’s product team in this wave of AI innovation.

Simultaneously, observers have noted that the Meitu team also adopts a well-considered approach to implementing AI at the product feature level.

For instance, in product design, Meitu’s team not only relies on data insights but also assembles a team of designers and fashion editors to continuously monitor global visual trends on social media. This enables them to swiftly translate popular makeup looks, filter styles, and photographic compositions into models that AI algorithms can understand and replicate.

Some hit features are particularly attentive to addressing user pain points and are designed for simplicity and ease of use.

For example, the “AI Flash” feature garnered 1.9 million global users in its first month. A review of its launch process reveals that it addresses the question of why ordinary users should embrace AI. Common photography issues like “backlit and dark faces” can be resolved by AI simulating professional lighting effects. From a product interface perspective, the operation is straightforward enough that even novice users can solve the problem with a single click.

Meitu CFO Yan Jinliang emphasized the commercial essence of “core feature compounding” in an interview: “Users rarely pay for short-term gimmicks but are willing to pay for core features that can be used repeatedly.”

To a certain extent, it is this user-centric perspective that has driven Meitu’s imaging and design product revenue to grow by 45.2% year-on-year, with the number of paid users surging more than eightfold in five years.

However, observers also believe that this “feature-driven” growth model harbors vulnerabilities.

First, Meitu’s core competitive edge lies in its long-accumulated aesthetic data, its understanding of users’ “beauty enhancement” needs, and its ability to translate trends into AI algorithms. However, the technological barriers of individual hit features are not insurmountable. The market is replete with various similar products, including those from industry giants, whose technological prowess and ecological synergy capabilities are formidable.

For example, products like “Xingtu” under ByteDance boast significant technological strength and can leverage seamless user acquisition and synergy from super apps like Douyin.

This means that Meitu needs to maintain speed and agility at both the product feature and market action levels to retain its core user base.

Second, paid penetration still requires improvement. Despite the rapid growth in the absolute number of paid users, as of mid-2025, its paid penetration rate is approximately 5.5%, still lagging behind international leading products like Canva, which have a paid penetration rate of around 10%. This indicates that the path to converting massive free users into paid users remains lengthy and is a key battleground for competing with rivals for paid users.

Meitu’s growth is highly contingent on whether its product team can continue to maintain the acuity and efficiency of a “hit-making machine.” However, it must be noted that creativity itself is not a readily replicable or predictable business. Translating occasional flashes of inspiration into product features that drive actual commercial success requires sustained and profound product insight capabilities. This is also a challenge Meitu must confront.

03

Betting on the AI Productivity Tools Market

Beyond the consumer (C-end) market, this wave of AI innovation has also prompted the company to set its sights on the productivity tools market, which offers even broader potential.

A landmark event was its deep strategic cooperation with Alibaba. In May of this year, Meitu and Alibaba announced a strategic partnership in which Alibaba would invest 250 million USD in Meitu through convertible debt. The two sides would engage in in-depth cooperation in e-commerce, AI technology, and cloud computing.

Specific cooperation terms include Alibaba promoting Meitu’s AI e-commerce tools on its e-commerce platforms and providing AI computing power support. In return, Meitu committed to purchasing no less than 560 million yuan worth of cloud services from Alibaba over three years. Additionally, Meitu would cease independent research and development of video foundation models and instead adopt Alibaba’s Tongyi multimodal and Wanxiang video large models as its base, conducting vertical training on top of them.

This cooperation has been interpreted by the outside world as a manifestation of Meitu’s mature “model container” strategy: Meitu no longer seeks to compete with tech giants in the race for foundational large models but instead positions itself as a top-tier “model application craftsman.” Its self-developed visual large model, MiracleVision (Qixiang Intelligence), focuses on aesthetic enhancement and has evolved to version 5.0, capable of generating high-quality videos. More resource-intensive general capabilities like video generation are supported by Alibaba Cloud’s computing power and Tongyi’s base models.

This strategy allows Meitu to “travel light” and focus its main resources on productization and commercialization.

The direct results of this cooperation have also enabled Meitu’s AI to have more implementations in ToB (business-to-business) productivity scenarios: Meitu Design Studio has been integrated into Alibaba’s Qianniu backend, with its “AI Product Image” feature directly serving millions of Taobao and Tmall merchants. Meitu’s traditional strengths in image processing can be translated into actual productivity for the e-commerce industry. Additionally, in July of this year, Meitu launched its AI Agent product “RoboNeo,” which integrates functions like image retouching, brand design, and web page creation. It supports completing complex imaging needs with “single-sentence instructions,” covering multiple scenarios and lowering the barrier to visual content creation.

Previously, Meitu’s division between ToB and ToC (consumer) segments was largely based on imaging and design products for consumers, with ToB revenue mainly coming from beauty industry solutions (SaaS), which were also defined as Meitu’s “non-core business.” In this year’s interim report, revenue from this segment declined significantly as Meitu focused on its main business and scaled back supply chain management services.

Now, the traditional ToC imaging and design segment is also expanding its coverage of ToB productivity tool scenarios, which is expected to become a new growth driver for Meitu in the future.

Financial reports show that as of June 30, 2025, Meitu had 23 million monthly active users (MAUs) in productivity scenarios, marking a 21.1% year-on-year increase, while the total MAU growth rate for lifestyle application scenarios was 7.5%. The user growth in productivity scenarios is evident.

Wu Xinhong also emphasized the immense potential of productivity scenarios during the interim financial report, believing that in the future, revenue and profits from productivity scenarios could account for more than half of Meitu’s totals. “That day may not be far off. Global productivity tools are still experiencing rapid growth,” he said.

04

Can It Maintain a Faster Pace in the AI Marathon?

Meitu, after its AI “awakening,” also needs to navigate a complex competitive environment.

Founder Wu Xinhong previously mentioned Meitu’s sense of crisis during a financial report conference. “It can be observed that many competitive AI applications come from startup teams rather than large corporations,” Wu said. “Startups have fewer burdens and no rigid frameworks, while large corporations often struggle to change course quickly due to their size.”

This reflects the current reality of the market. In fact, mid-sized companies are currently under dual pressure in this wave of innovation. On one hand, foundational model vendors and tech giants are rapidly moving into the tool application market. Major foundational model platforms and giants have all built powerful image generation and editing capabilities. ByteDance’s “Xingtu” also leverages its strong algorithms and traffic advantages to compete head-on with Meitu in feature iteration.

On the other hand, countless small startups are using open-source AI models to focus on niche feature points (such as AI portraits in specific styles or ID photo beautification) for single-point breakthroughs. They have extremely short decision-making chains and very low trial-and-error costs.

This forces Meitu into a dilemma: It enjoys high gross margins but must continuously invest significant marketing expenses to build its brand, lead trends, and maintain market presence. As competition intensifies, rising marketing expenditures are likely to erode its currently impressive profits.

Moreover, organizational agility presents a structural hurdle for this publicly - traded enterprise, which has endured setbacks and is now experiencing a resurgence. Internally, Wu Xinhong champions the concept of "one person acting as a team" as a means to combat corporate red - tape and rekindle the entrepreneurial zeal. Yet, preserving the concentrated focus and nimbleness of a startup within a system comprising nearly 1,300 employees is an arduous and protracted struggle.

Wu Xinhong once drew an analogy between the company's strategy and auto racing, stating, "Winning isn't merely about slamming on the gas pedal; the way you navigate the curves, entering and exiting them, holds greater significance."

In the initial curve of the AI race, Meitu adopted an "applications - first, lightweight overtaking" strategy and managed to secure a temporary lead.

Now, the path ahead is even more convoluted. The company must strike a balance between relying on upstream model providers and constructing a robust downstream competitive moat. It needs to bridge the chasm between creating short - lived, popular features and establishing a long - lasting product ecosystem. Additionally, it must find an equilibrium between driving profit growth and navigating through fierce competition.

Meitu's transformation narrative serves as a case study in intelligent survival amidst resource and capability limitations. Its "rebirth" is indeed remarkable. However, this race against time, competitors, and its own inherent characteristics may still have its crucial turning points waiting to be encountered further down the road.