For eighteen years, Meitu has been the definitive face of China’s digital aesthetic, evolving from a simple PC photo-editor into a mobile juggernaut that defined the 'selfie culture' of the 2010s. However, as 2026 unfolds, the company finds itself at a precarious crossroads. The rise of Large Language Models (LLMs) and autonomous AI Agents threatens to 'devour' the traditional software model, rendering the manual, step-by-step editing process of the past obsolete.
In an exclusive dialogue, Meitu Chairman Wu Xinhong reveals a radical strategy to survive this disruption: tearing down the company’s walled garden. By opening eight core AI imaging capabilities to the 'OpenClaw' ecosystem, Meitu is transforming its proprietary tools into shared infrastructure. This move signifies a shift from being a destination app to becoming the underlying 'plumbing' for a new generation of third-party AI developers and competitors.
This 'skills-open' approach is a calculated response to the changing nature of human-computer interaction. Wu notes that the industry is moving from mouse clicks and touchscreens to natural language interfaces where a single command can replace minutes of manual labor. If a user can simply tell an AI Agent to 'beautify these 100 photos according to my preference,' the traditional app interface loses its primary reason for existing.
Meitu’s pivot also reflects the brutal financial reality of the AI arms race. After a brief attempt at building its own general LLM, the company pivoted to a 'model container' strategy, utilizing a mix of vertical proprietary models and third-party APIs. The decision was driven by the realization that an application-focused company cannot sustain the multi-billion dollar capital requirements of a foundational model war against giants like ByteDance or Alibaba.
Internally, the company has restructured into a venture-capital-style incubator to foster rapid innovation. Seven AI-focused studios have been seeded with 10 million RMB each, while hundreds of internal teams compete in 'survival of the fittest' tournaments. This bottom-up approach acknowledges that in a period of such rapid technological shifts, management can no longer predict the 'winning' product form by fiat.
Despite these efforts, the transition is not without friction. Meitu’s gross margins have seen a slight contraction as the costs of third-party API calls mount. The company’s long-term survival depends on its ability to move users from established legacy apps to native AI Agents while simultaneously proving that specialized, high-fidelity imaging remains a niche that generic models cannot yet perfect.