--- title: "Volume surged by 8%! Tencent released an AI image model, benchmarking against the popular logic of Meta Muse" type: "News" locale: "en" url: "https://longbridge.com/en/news/299720317.md" description: "Tencent's Hong Kong stock surged nearly 8%, primarily due to the popularity of Meta's AI Agent application Muse, which triggered a market mapping logic: WeChat, with its vast social and mini-program ecosystem, is seen as the most advantageous landing platform for the \"Chinese version of Muse.\" The market is optimistic about its ability to leverage mini-programs and a closed-loop payment system to reduce AI delivery friction, seizing the first-mover advantage in the 2C Agent track through distribution and ecological advantages" datetime: "2026-09-22T06:44:38.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/299720317.md) - [en](https://longbridge.com/en/news/299720317.md) - [zh-HK](https://longbridge.com/zh-HK/news/299720317.md) generator: "portal-rs" --- # Volume surged by 8%! Tencent released an AI image model, benchmarking against the popular logic of Meta Muse Tencent Holdings saw its Hong Kong stock rise nearly 8% on Tuesday, with trading volume reaching HKD 17.7 billion. The catalyst came from two interrelated news lines: Meta's AI Agent application Muse quickly gained popularity in the U.S., topping the app store download charts, while Tencent simultaneously released its latest image generation model, Hy Image 3.5 Preview. This led the market to form a clear mapping logic—Tencent, with its super social ecosystem and high-frequency life scenario entry points, is viewed as the most qualified carrier for the "Chinese version of Muse." The day after Meta Muse launched, it entered the top three of the U.S. App Store's free chart and maintained a leading position in the productivity rankings. The core difference of Muse lies in its ability to transcend the chatbot category: by reading social relationships and accessing various APIs such as email, calendar, shopping, and food ordering, it autonomously executes complex cross-application tasks for users in the background. This blockbuster effect has prompted the capital market to turn its attention to domestic comparable companies. With its vast mini-program service ecosystem and closed-loop payment system, WeChat is believed to have the potential for efficiency and commercial monetization that rivals Meta once AI Agents are deeply integrated, making Tencent the most sought-after beneficiary company of the day. CICC's research report pointed out that the breakthrough for Meta Muse has shifted from "Is the model smart enough?" to "Can users trust it, and can the product connect to real-world tasks?" The competitive focus of 2C Agents has officially shifted from the technical level to ecological distribution and trust architecture. Gary Tan, a portfolio manager at Allspring Global Investments, stated, "Some investors are comparing Tencent to Meta, especially considering WeChat's unique social ecosystem and its potential to support large-scale personal AI assistants." ## Tencent Releases Image Model on the Same Day, Accelerating Its Pace Tencent's release of Hy Image 3.5 Preview is its latest move to compete directly in the AI image generation track. Tencent stated in a press release that this model has improved performance compared to its predecessor and is now integrated into its Yuanbao and video editing, design tool product lines. Tencent has used hundreds of internal designers as a testing group and stated that the model's performance is comparable to ByteDance's Seedream 5.0 Pro, while slightly better than Alphabet's Google Nano Banana Pro and Alibaba's Qwen-Image-3.0 Pro. Tencent has thousands of game designers and artists, who are expected to be the main beneficiaries of this tool; however, Tencent did not quantify the quality claims mentioned aboveThe release coincides with Alibaba's launch of the artificial intelligence conference, making the timing quite striking. Since hiring former OpenAI researcher Yao Shunyu as Chief AI Scientist, Tencent's AI strategy has shifted: Yao has publicly criticized the practice of training models solely for leaderboard scores, emphasizing product integration and solving real-world problems. Another former OpenAI researcher and computer vision expert, Tian Yonglong, also joined Tencent's Hunyuan team in July to lead the development of visual language models. ## Muse's Hit Logic: A Paradigm Shift from Chatting to Delivery To understand the core of Tencent's recent surge, it is essential to clarify what new narrative Meta Muse has opened up. CICC's research report outlines four substantial differences between Muse and previous similar Agent products: **First,** asynchronous execution in the background retains long-term memory and user preferences across conversations, allowing tasks to continue even after the user closes the application; **Second,** real account connections through three methods: built-in connectors, public APIs, and browser operations, integrating scenarios such as email, calendar, payments, health, e-commerce, and smart home; **Third,** isolation of execution and approval gating, where each user has an independent Muse Secure VM runtime environment, guarded by Sentinel for all external operations, making security architecture a core aspect of product design; **Fourth,** the Ideas and Feed functions have revealed the embryonic form of a recommendation-based Agent, with Muse beginning to shift from waiting for user questions to proactively identifying needs and defining problems in advance. CICC's report believes that the competition for 2C Agents will resemble "recommendation" rather than "search"—Agents will actively identify needs based on sufficient context, defining and solving problems for users instead of passively waiting for instructions. Meta's advantage lies in its distribution and trust infrastructure: 3.6 billion daily active users provide a scale for reach, WhatsApp conversation threads lower the usage threshold, and the long-term content behavior accumulation on Instagram and Facebook forms a richer personal demand map, which is difficult for products like ChatGPT that rely solely on chat history to replicate. In terms of business model, Muse adopts a free tier and two monthly subscription levels at $20 and $100, with the highest tier pricing approaching that of some ToB software. CICC's report suggests that as Muse takes on shopping, booking, and other consumer tasks, its monetization path is expected to form a hybrid model of "subscription base + transaction commission." ## Why Tencent is the Optimal Solution for "China's Version of Muse" The capital market's comparison of Tencent to Meta Muse is not a simple analogy; it is supported by a clear ecological logic. CICC's research report points out the differences between the 2C Agent ecosystems in China and the U.S., stating that **WeChat has integrated accounts, payments, and lifestyle services within its platform, and possesses a unique mini-program ecosystem that facilitates its own Agents to directly invoke capabilities and complete transactions within the authorized scope.** In contrast, Meta's overseas service entry points are relatively dispersed, with web pages still being an important carrier. Muse needs to rely on browser operations to cover long-tail services, resulting in relatively high connection and maintenance costs, and web page revisions may also affect execution efficiency. In other words, if Tencent deeply integrates AI Agents into the WeChat ecosystem, its natural mini-program closed loop and payment system will provide it with lower execution friction compared to Meta. The vision Tencent is striving to achieve—building an AI capable of executing various tasks for over 1 billion users in the WeChat ecosystem—aligns closely with Muse's product direction, which is the core logic behind the market's excessive pricing of Tencent. ## Timing Determines Rhythm: Two Paths and Four Stages CICC's research report introduces a product innovation cycle framework, dividing the development of 2C Agents into four stages: exploration phase (where both technical effects and costs have not reached the threshold), transitional innovation phase (where technical effects reach the threshold but costs remain high), all-encompassing innovation phase (where both effects and costs break through the threshold), and strong get stronger phase (where the technology curve flattens). The report uses the historical paths of Blackberry and iPhone as a reference: the transitional innovation phase corresponds to Blackberry's focus on enterprise email and reduction of non-core functions, with Blackberry's revenue increasing from $600 million in 2004 to a peak of $20 billion in 2011; the all-encompassing innovation phase corresponds to the iPhone's all-in-one integration, which ecologically crushes transitional products. CICC's research report believes that the continuous improvement of AI model capabilities and the decline in Token computing costs are currently highly certain factors; what is more difficult to grasp is users' tolerance for errors in Agent task solutions, their willingness to migrate from existing habits, and the user cost threshold. Based on this, there are two development scenarios for 2C Agents: one is to first experience the transitional innovation phase, starting from general office needs and gradually extending to comprehensive scenarios; the other is that the technology and cost curves decline rapidly in sync, directly crossing the transitional phase into the all-encompassing innovation phase—at that time, giants like Meta and Tencent, which possess distribution advantages and ecological depth, will benefit first. CICC emphasizes that the shorter the interval from the transitional phase to the all-encompassing phase, the greater the advantage for internet giants; conversely, startups will have more time to build barriers ### Related Stocks - [TCTZF.US](https://longbridge.com/en/quote/TCTZF.US.md) - [TCEHY.US](https://longbridge.com/en/quote/TCEHY.US.md) - [00700.HK](https://longbridge.com/en/quote/00700.HK.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [FBL.US](https://longbridge.com/en/quote/FBL.US.md) - [FBY.US](https://longbridge.com/en/quote/FBY.US.md) - [FBYY.US](https://longbridge.com/en/quote/FBYY.US.md) - [METD.US](https://longbridge.com/en/quote/METD.US.md) - [METU.US](https://longbridge.com/en/quote/METU.US.md) - [METW.US](https://longbridge.com/en/quote/METW.US.md) - [MAGX.US](https://longbridge.com/en/quote/MAGX.US.md) - [03908.HK](https://longbridge.com/en/quote/03908.HK.md) - [601995.CN](https://longbridge.com/en/quote/601995.CN.md) - [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md) - [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md) - [09988.HK](https://longbridge.com/en/quote/09988.HK.md) - [BABA.US](https://longbridge.com/en/quote/BABA.US.md) - [OpenAI.NA](https://longbridge.com/en/quote/OpenAI.NA.md) - [80700.HK](https://longbridge.com/en/quote/80700.HK.md) - [HTCD.SG](https://longbridge.com/en/quote/HTCD.SG.md) - [89988.HK](https://longbridge.com/en/quote/89988.HK.md) - [HBBD.SG](https://longbridge.com/en/quote/HBBD.SG.md) ## Related News & Research - [Meta tested call-center contractors behind Muse AI calls for half its staff](https://longbridge.com/en/news/299918653.md) - [Meta Oversight Board Says Meta's Safeguards Inadequate To Address AI-Generated Content On Its Platforms](https://longbridge.com/en/news/299301748.md) - [Manus seeks $4B valuation in new $500M fundraise as it resumes independent ops](https://longbridge.com/en/news/299472778.md) - [Meta’s Muse hits Mac, letting the AI take actions on your computer](https://longbridge.com/en/news/299467411.md) - [Amazon blocks Metaâs Muse AI agent from its retail site](https://longbridge.com/en/news/299735986.md) --- > **Disclaimer: This article is for reference only and does not constitute any investment advice.**