Oppenheimer's math reveals the steep hill Meta must climb to make its new AI agent a meaningful revenue driver beyond advertising
Meta’s newly launched AI agent, Muse, would need roughly 115 million paying subscribers at $20 per month to meaningfully move the needle on the company’s earnings, according to Oppenheimer analyst Jason Helfstein. That would translate to approximately $27.5 billion in annual revenue, assuming an 80% operating margin. For context, that subscriber count would be larger than the entire population of Japan.
The trust gap and the competition problem
Muse, which debuted on September 8, 2026, positions itself as a personal AI agent that can execute tasks across applications like email and travel bookings. To do that effectively, though, it needs access to your accounts, passwords, and personal data.
Only 13% of surveyed US consumers said they’d be willing to share all their passwords with Meta. Then there’s the competitive landscape. ChatGPT and Gemini already have substantial footholds in the AI assistant market.
Muse operates on a freemium model with three tiers. The free version offers 100 million tokens per week. The Power plan runs $20 per month with 500 million tokens. And the Maximum plan costs $100 per month, providing 3 billion tokens plus priority access.
Wall Street’s more moderate expectations
Goldman Sachs recently assigned Meta a Buy rating, projecting that even 10 million subscribers could generate approximately $2.4 billion in annual revenue from the service.
Wedbush took an even more bullish stance, raising its Meta price target to $650 and estimating that Muse subscription revenue could reach about $5 billion by 2027. Even at that level, it would represent only about 1.5% of Meta’s anticipated total revenue for that year.
Why Meta is chasing subscriptions at all
As of Q2 2026, Meta reported revenue of approximately $60.8 billion, with advertising contributing around 98% of that total.
Muse represents the most direct attempt yet to monetize Meta’s AI investments through consumer-facing products. The freemium model also serves a secondary purpose: data collection. Even users who never pay for Muse will interact with it, generating training data and usage patterns that could improve the product over time.
