Meta has officially revealed during its third-quarter earnings call that its dedicated AI application, Muse, has crossed the milestone of 10 million users, validating earlier market intelligence estimates.
While rapid user acquisition is a significant metric, Meta CEO Mark Zuckerberg emphasized that user engagement patterns, transaction volumes, and financial behaviors within the application present an even more compelling narrative for the future of agentic AI.
Strong Engagement Metrics and Agentic Shopping Dynamics
Data disclosed during the Q3 earnings presentation demonstrates that Muse is generating impressive retention figures. Approximately 25% of users who set up the Muse application transition into daily active users. A primary driver of this engagement stems from agentic e-commerce capabilities built directly into the platform.
Zuckerberg highlighted that over a quarter of Muse's engaged users who utilize the AI agent for shopping generate an average transaction size exceeding $200. This level of high-value commerce activity exceeded early industry expectations and signals a major shift in how consumers utilize artificial intelligence for purchasing decisions.
“Plaid told us that they hadn’t seen activity like they’re seeing with their connector on any other service before,” stated Meta CEO Mark Zuckerberg, highlighting the financial technology underlying Muse's shopping workflows.
To facilitate these seamless agentic transactions, Muse leverages financial technology partner Plaid to connect users' bank accounts and payment applications directly to the AI agent. This integration allows Muse to handle commercial workflows end-to-end on behalf of the user.
Productivity Workflows and Virtual Machine Security Infrastructure
Beyond agentic shopping, personal productivity tasks represent another dominant use case for Muse. Users are routinely granting the AI agent permission to interface with critical personal tools, including email accounts and digital calendars.
Meta attributes this deep integration to user confidence in the platform's specialized security architecture. The system currently operates on an infrastructure layer known as Muse Secure VM, which is engineered to shield user data and maintain strict privacy boundaries.
Privacy Roadmap: Muse Confidential VM
To further bolster security for sensitive user workflows, Meta is preparing to launch Muse Confidential VM. This upcoming architectural enhancement is designed to restrict access so completely that even Meta itself will be unable to view or access user content processed within the environment.
Monetization Model and Compute Optimization Strategy
With Muse establishing clear product-market fit, Meta is shifting its primary focus toward scaling the user base, driving revenue generation, and optimizing compute expenditure. Meta forecasts that Muse will ultimately produce higher average revenue per user (ARPU) than its core family of apps—which currently includes Facebook, Instagram, WhatsApp, and Messenger, yielding several hundreds of dollars annually per user.
Meta plans to monetize the Muse platform through a multi-tiered commercial strategy:
- Transaction Commission: Keeping the core application free for general users while retaining a percentage cut of commercial transactions executed on the platform.
- Power User Subscriptions: Charging power users who require access to Meta's most advanced AI models or higher token capacities.
Although Muse is currently compute-intensive to run, Meta is expanding infrastructure capacity to serve more than one billion users over the coming years. The company projects a greater than 10x improvement in compute cost efficiency within the next year, paving the path toward profitability once revenue streams reach scale.
Hardware Ecosystem and the Future 'Muse Network'
Muse is set to serve as a central software component within Meta's broader hardware strategy, particularly for smart glasses managed by the Reality Labs division. While Reality Labs continues to operate at a loss, financial metrics are improving, with projected division losses decreasing from $19.1 billion in 2025 to $15 billion this year.
Looking ahead, Zuckerberg outlined a broader conceptual vision referred to as the Muse Network. This strategy envisions transitioning Muse from an isolated, single-user tool into an interconnected ecosystem.
Under the Muse Network framework, individual user agents will interact autonomously with the AI agents of friends, family, and commercial entities. Rather than navigating traditional websites to make purchases, users' personal Muses will communicate directly with business-owned AI agents to negotiate and execute transactions seamlessly.