Released in 23 days, it has become a sensation in the consumer agent market! Morgan Stanley provides an in-depth analysis of the "Muse 6 Key Questions".

Meta's AI agent application Muse quickly ignited the consumer agent market in just 23 days after its launch, becoming one of the fastest-growing AI applications recently.

According to data from market research firm Sensor Tower, as of September 30, Muse had surpassed 5 million downloads. Muse launched in the US and Canada on September 8 and reached this milestone in just 22 days; in comparison, ChatGPT, Grok, and Claude took 56 days, 103 days, and 492 days, respectively, to reach 5 million downloads.

Muse's rapid growth has drawn market attention to its costs, computing power, and monetization capabilities . A recent in-depth study by Brian Nowak, chief analyst for the internet sector at Morgan Stanley, and his team analyzed six key areas: Muse's monetization, service costs, computing power reserves, enterprise market, Amazon partnership, and valuation .

The core contradiction lies in the fact that the faster the Agent user growth, the higher the computing power cost; while Meta needs to prove that future advertising, transactions, enterprise subscriptions and API businesses can cover this investment and convert the huge user base into revenue.

I. Monetization: In the short term, focus on expanding the user base; in the long term, target the $30 trillion consumer market.

Muse is not in a hurry to make money at the moment. Morgan Stanley expects that subscription revenue will contribute limitedly in the short term, and transaction commissions will remain at a low level. Meta's more important task is to expand its user base, increase usage frequency, and cultivate users' habit of completing commercial transactions through agents.

Large-scale monetization may not occur until 2027-2028 . In the long term, Meta may build an agent-based commercial market around real-time bidding: companies bid based on their expected rate of return and pay transaction fees to Meta.

This model corresponds to a potential market size of $30 trillion , covering consumer scenarios such as retail, travel, transportation, food delivery, advertising, logistics, and wearable devices.

Before charging directly, Muse can first enhance the monetization efficiency of its Meta advertising business by improving user engagement and user profiling capabilities.

II. Cost: Up to $130 per user per month; high cost may actually become a barrier to entry.

The more popular Muse becomes, the more Meta faces the issue of cost.

Morgan Stanley estimates that the cost of the Muse service is approximately $3 to $130 per user per month, primarily depending on token consumption. GPU inference costs increase with token usage, while CPU sandbox costs depend on user activation frequency and virtual machine runtime.

Based on a user consuming 25% of their weekly token quota, the cost per user per month is approximately $37. Since this calculation assumes Meta rents all CPU sandbox capacity at AWS on-demand pricing, the actual cost may be lower.

This also means that Agent is not a business that can be easily replicated by burning money. Meta has a core advertising business of approximately $250 billion, growing at a rate of 27% year-on-year, as financial support, and an average revenue per user of about $70 globally, which can support the high upfront investment in AI infrastructure.

III. Computing Power: Theoretically, CPU procurement in 2026 is sufficient to support 1 billion DAU.

If Muse continues to grow rapidly, will Meta have enough computing power?

Morgan Stanley believes that Meta's current CPU procurement scale is theoretically sufficient to support 1 billion daily active users of Muse.

According to its estimates, Meta will purchase a total of approximately 643 million vCPUs in 2026, corresponding to approximately 321 million Muse instances. Under the assumptions of 3 hours of daily activity, double the peak factor, and 1.2 times the capacity buffer, approximately 300 million concurrent active instances are needed to support 1 billion DAU, which can be largely covered by existing computing power.

This means that as Muse's user base expands, Meta's real short-term challenge may not be "whether it has computing power," but rather how to convert its massive computing power investment into user growth and business revenue.

The report projects that Meta's capital expenditures in 2027 will be approximately $225 billion, with corresponding depreciation of approximately $59 billion; third-party computing contract expenditures will be approximately $26 billion, up from $17 billion in 2026.

IV. Enterprise Market: Entering the $25 Trillion Knowledge Economy from the Consumer Agent Perspective

Muse's commercial vision extends beyond individual users.

Meta is pushing its Agent further into the enterprise market, targeting the approximately $25 trillion global knowledge-based work economy. Related products include Muse for SMEs, WhatsApp and Messenger Agent tools, enterprise products, Muse Coding, and the future model API "Watermelon".

Morgan Stanley believes that the traditional Web 1.0 architecture of enterprises is not suitable for direct interaction between agents. In the future, enterprises will need to provide machine-native interfaces that allow agents to directly access products, services, and enterprise systems.

This provides Meta with a new pricing model: charging enterprises subscription fees while providing API computing power and model services to developers and enterprises.

Sensitivity analysis shows that if 60 million users pay $15 per month, it could increase EPS by about 8.2% by 2028; if 100MW of dedicated computing power is used for API business, it could also increase EPS by about 4% by 2028.

V. Amazon: With the entry of agents into e-commerce, the transaction portal becomes the core of the game.

Once Muse starts shopping for users, Meta will inevitably have to deal with Amazon.

Amazon holds approximately 40% of the US e-commerce market and controls a vast system of goods, merchants, and logistics. Key issues surrounding potential collaboration between the two companies include who records merchant activity, who controls the shopping experience, how Prime benefits and retail media value are distributed, and which data can be fed back to Muse.

At the same time, Meta relies on Amazon's AWS cloud resources, including CPU and GPU computing power, as well as distribution channels that provide model APIs through Bedrock.

Therefore, the potential collaboration between Meta and Amazon is not just a typical e-commerce partnership, but may also involve transaction access, user data, cloud computing power, and model distribution.

VI. Valuation: The PE ratio of 21x has already reflected some of the expectations for Muse; the next step is to focus on commercialization.

Following the release of Muse and the settlements reached between Meta and state attorneys general, Meta's NTM consensus EPS valuation has risen from approximately 14 times to 21 times.

Morgan Stanley believes that the current valuation already reflects some of the first-mover advantage gained by Meta's free launch of Muse. Whether the valuation can be further increased depends on three key factors: first, whether Muse's user base and usage rate can continue to grow; second, whether the Agent can truly generate large-scale commercial transactions; and third, the pace of development for competitors such as Google and SPCX.

Meta reached a valuation of 28 times in February 2025. Therefore, if Muse can continue to expand its user base and further demonstrate its commercial potential, the market may still reassess its valuation level.

From a fundamental perspective, Morgan Stanley focuses on the following upside drivers: AI-driven increases in core user engagement, growth in business messaging and agent businesses, and the gradual scaling of API platform monetization.

Ultimately, Muse's most important value right now isn't how much revenue it contributes, but whether Meta can leverage its massive user base, advertising cash flow, and computing power investment to transform a high-cost AI application into an agent platform connecting consumers, businesses, and commercial transactions.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.