The explosive popularity of Google's new AI model Nano Banana has sparked profound market concerns about whether "AI will devour application software." However, Morgan Stanley analysts believe such worries are excessive.

On September 2, Morgan Stanley's research report indicated that Meitu's growth trajectory remains unaffected by the AI model Nano Banana, with the company's true value lying in providing "last mile" solutions that foundational AI models cannot achieve.

The report emphasizes that Meitu has built solid competitive moats through its deep accumulation in vertical domains, proprietary data, and focus on core paid functionalities. Additionally, the company maintains openness and neutrality toward all AI models and has seamlessly integrated new technologies like Nano Banana into its products.

Morgan Stanley reaffirmed its "Overweight" rating on Meitu with a target price of HK$15.70, believing its long-term value remains solid. Revenue is projected to reach RMB 4.021 billion in 2025, growing to RMB 5.120 billion in 2026, and further expanding to RMB 6.771 billion in 2027.

**Market Panic Triggered by Nano Banana**

Recently, the new image AI model "Nano Banana" released by Google rapidly gained popularity on Chinese social media platforms.

Released on August 26, the model ignited public discussion on August 30 with its powerful capability to transform photos into 3D model images through prompt words.

This event immediately triggered investor concerns that increasingly powerful foundational AI models would cross boundaries to "erode" the core businesses of application software companies like Meitu.

Affected by this panic sentiment, Meitu's share price plummeted 14% during the trading day when the news fermented, while the Hang Seng Index rose 2% over the same period.

Notably, this is not an isolated incident.

Throughout this year, whenever major AI models like GPT-4o or Google's Veo3 have released updates, Meitu's stock price has experienced similar fluctuations.

This reflects widespread market anxiety about the survival model of application-layer software in the AI era.

**"Last Mile" Value Difficult to Replace**

Morgan Stanley analysts emphasize that in the AI era, the value of application software lies in providing "last mile" services to optimize results, which is precisely what general-purpose AI models cannot fully achieve.

The "last mile" in the visual industry is particularly complex for two reasons:

**Scenario fragmentation and demand diversity**: User needs vary dramatically, from personal photo enhancement and advertisement generation to poster design. Each scenario has unique standards and workflows.

**Subjectivity of personal preferences**: Aesthetic and visual preferences are highly personalized, making them difficult to capture accurately through standardized prompt words and impossible for general-purpose AI models to implement perfectly.

These two characteristics determine that general-purpose AI models struggle to provide end-to-end perfect products meeting all user needs.

Therefore, the existence value of application software focused on specific scenarios and optimizing user experience has not been weakened but has become increasingly prominent.

**Meitu's Advantages: Scenarios, Data, and Core Payment Points**

Morgan Stanley believes Meitu has established significant advantages in maximizing "last mile" value through its long-term accumulation.

First, the company maintains business focus and has accumulated deep scenario understanding. Meitu concentrates on non-professional casual and non-professional productivity scenarios, further refining the latter into areas like e-commerce design.

This focus enables deep understanding of user pain points in specific scenarios and development of highly optimized workflows that enhance user experience.

Second, massive vertical data creates barriers. Through partnerships with companies like Alibaba and long-term operation of its own products, Meitu has accumulated massive, high-quality vertical data, particularly in portrait beautification and e-commerce design.

Meitu leverages this proprietary data to conduct "post-training" on third-party foundational large models, creating small models with effects far superior to general-purpose models and results more aligned with user needs.

Finally, a clear business model: users pay for "core functions." Meitu's functionalities can be divided into "entertainment features" (like AI emoji generation) and "core functions" (like AI facial refinement).

The former may attract users to try the product, but the latter is the fundamental driver of user paid subscriptions. General-purpose AI models can mimic the former (like Nano Banana's 3D models) but struggle to penetrate the high-quality "core function" domains provided by Meitu.

Additionally, Meitu adopts an AI model-neutral strategy, utilizing AI model container technology to rapidly integrate capabilities from different AI models into its products.

Before 3D model images gained widespread circulation, Meitu had already integrated Google's Nano Banana into its overseas products, including Meitu Xiuxiu, BeautyPlus, and RoboNeo.