微信AI入口已就位 抖音跟不跟？

This article outlines the latest AI developments of China's two top super apps, WeChat and Douyin. Key takeaways are as follows:

1. WeChat is currently conducting a small-scale internal test of its native system-level AI assistant, Xiaowei. Placed at the top-left corner of WeChat's main interface, Xiaowei runs on WeChat's self-developed WeLM large language model and also integrates high-quality external models. The feature currently cannot be turned off. Xiaowei supports operating native WeChat functions, launching mini-programs for various lifestyle services, searching and generating content, and generating mini-program prototypes from a single sentence. Currently, generated mini-programs are only accessible to their creators.

2. Douyin has not yet launched a unified AI entry point. It has adopted an approach of embedding AI functions across scattered scenarios paired with collaboration with external AI chatbot Doubao. AI functions have already been rolled out in social interaction, lifestyle services and content creation. Douyin has also fully connected Doubao with Douyin E-commerce end-to-end, allowing users to complete the entire transaction process from product selection to after-sales service directly within Doubao.

This article discloses the latest AI strategies of China's two largest super platforms, WeChat and Douyin, offering actionable insights for brands to navigate channel shifts and capture AI-driven traffic opportunities. Key takeaways are as follows:

1. Built on WeChat's 1.432 billion combined monthly active users (MAUs) and millions of existing mini-programs, the new unified system-level AI entry Xiaowei will soon be able to recommend brand services and launch brand mini-programs for transactions directly via AI conversation, significantly shortening the consumer journey. Brands that adapt their mini-programs for WeChat's AI ecosystem early stand to gain access to a valuable new conversion channel.

2. Douyin has completed end-to-end integration between Doubao and Douyin E-commerce, enabling brands to reach consumers and close transactions directly through Doubao conversations. Douyin's AI travel assistant for local lifestyle services can also recommend brand merchants' group-buy offers and redirect users to checkout. Additionally, Douyin's "Little Fire Man" social feature has surpassed 100 million daily active users, and AI-powered social expansions have opened up new social marketing opportunities for brands.

This article summarizes the latest AI strategic layouts of WeChat and Douyin, China's two leading e-commerce transaction platforms, and highlights new growth opportunities for sellers. Key takeaways are as follows:

1. Opportunities on WeChat: WeChat has launched Xiaowei, a unified system-level AI entry that supports launching mini-programs for transactions directly and generating mini-program prototypes from a single user prompt. This allows small and medium-sized sellers to test their mini-program business at low cost, lowering the technical barrier to building an online channel. In the future, sellers can leverage WeChat's over 1.4 billion MAU base to acquire new customers driven by AI recommendations.

2. Opportunities on Douyin: Douyin has completed full-link integration between Doubao and Douyin e-commerce, allowing sellers to access Doubao's AI recommendation channel and gain new traffic. Local lifestyle sellers can capture recommendation traffic from Douyin's AI travel assistant to increase exposure. Since Douyin has not yet launched a unified AI entry, there are still open opportunity windows for future layout adjustments, and sellers can prepare early by adapting to the new ecosystem.

The latest AI layouts of China's two leading super platforms bring new insights and opportunities for factories to advance digital transformation and expand e-commerce sales channels. Key takeaways are as follows:

1. Business opportunities: WeChat's Xiaowei supports generating mini-program prototypes from a single sentence. When factories aim to build an online sales channel for their own brands, they can leverage AI to quickly generate mini-program prototypes at low cost and test market demand, drastically lowering the technical threshold and trial-and-error costs for factories to reach end consumers directly. Backed by WeChat's over 1.4 billion MAU base, factories' self-owned brands can gain greater incremental growth space.

2. Insights for digital transformation: Both platforms are accelerating the deployment of AI capabilities, and consumer purchase journeys will shorten further, with users able to complete product selection and checkout directly via AI conversation. Factories need to adapt to the new AI traffic rules in advance by organizing standardized product information to facilitate AI crawling and recommendation. They can also leverage AI tools to optimize product design and better align with consumer demand.

This article discloses the latest progress of AI entry layout in the super app sector, pointing out industry trends and market opportunities for AI service providers. Key takeaways are as follows:

1. Industry development trend: AI competition has evolved from the early battle for new users among standalone large model apps to a battle for AI central entry points among super apps that control massive existing user bases. Leading super platforms are accelerating the deployment of native AI capabilities and integrating their existing ecosystem resources, marking that AI adoption has entered a new phase of ecosystem competition.

2. Market opportunities and customer pain points: Leading platforms need to integrate capabilities from multiple large models to meet scenario-specific demands—for example, WeChat's Xiaowei uses both self-developed models and external open-source models, creating cooperation space for third-party large model service providers. In addition, platforms need to integrate AI capabilities across scattered scenarios to deliver a unified user experience, creating broad market demand for AI generation tools for ecosystem developers, such as AI-powered mini-program generation services.

The AI deployment practices of two leading super platforms, WeChat and Douyin, provide valuable references for AI upgrades for all types of platforms. Key takeaways are as follows:

1. The industry has already developed two mature AI adoption paths, which platforms of different types can reference: The first is WeChat's approach of a built-in unified system-level entry. Leveraging its massive existing MAU base and complete ecosystem, this model builds an all-scenario connected integrated AI assistant that connects all native functions and services to improve user experience and retention. This path is suitable for mature super platforms with fully developed ecosystems. The second is Douyin's approach of embedding AI across scattered scenarios paired with external large model collaboration. This approach distributes AI capabilities across vertical scenarios and partners with leading large models to connect ecosystems, leveraging the strengths of both parties to lower deployment costs.

2. Risk reminder: As user demand for an integrated intelligent AI experience continues to grow, platforms need to balance deployment costs and user experience, and avoid the risk of poor entry design that hurts user experience. At the same time, platforms can leverage AI capabilities to optimize merchant recruitment policies and attract more developers and merchants to join the platform.

This article summarizes the latest developments in China's large model industry deployment, demonstrates the characteristics of AI industry competition entering a new stage, and offers important reference value for research on AI transformation of the internet industry. Key takeaways are as follows:

1. New industry developments: AI competition has evolved from the early battle for user acquisition among standalone large model apps to a battle for AI central entry points among super apps with hundreds of millions of existing users. The focus of industry deployment has shifted from large model training itself to ecosystem integration and entry competition, and the industry has entered a new stage of competition.

2. In terms of business models, two distinct scalable AI adoption business models have emerged for research: The first is the WeChat model, which builds a unified built-in AI entry on top of its existing ecosystem, integrates all internal services to deliver an integrated intelligent experience, embeds AI capabilities as the core competitiveness of its own ecosystem, and expands the value of its proprietary ecosystem. The second is the Douyin model, which adopts open cooperation and division of labor, deploys AI capabilities across scattered scenarios, and partners with external large models to connect the transaction ecosystem, allowing all parties to gain benefits while lowering deployment costs. The impact of the future development of these two models on the industry landscape deserves in-depth research.