External estimates suggest scaling Muse could cost up to $84 billion

Meta has a new AI agent, and it is hungry. CFO Susan Li has flagged that computing costs for Muse, the company’s personal AI agent, are expected to climb as more people use it.

The counterweight is a bold efficiency target. Meta expects to improve compute cost efficiency by more than ten times over the next year.

What Muse is, and why it costs so much

Muse went live on September 8, 2026. It is pitched as a personal AI agent that carries out complex, multi-step tasks for users across various apps and websites.

That distinction matters for the cost story. A conventional chatbot answers a question and stops. An agent keeps working: it plans, clicks, checks, and tries again until the job is done.

Li’s message centers on exactly that dynamic. As the user base grows, the infrastructure needed to keep all those agents running grows with it. Meta’s plan to offset the pressure relies on better utilization of its hardware and optimizations to the models themselves.

The numbers outside Meta are doing

External estimates suggest serving 100 million Muse users would require approximately 65,000 CPUs and 75 PB of DRAM. That works out to around $2.8 billion in costs, according to those estimates.

Scaling Muse to Facebook’s estimated 3 billion users could cost approximately $84 billion, according to unverified external calculations. Those figures are not Meta’s, and they assume today’s cost structure holds, which is precisely what Meta says it plans to change.

Meta is not leaving everything to engineering, either. Paid subscription tiers for Muse are priced at $20 or $100 per month, aimed at covering the costs of heavy users. The free tier comes with limitations.

Capex keeps climbing

Muse arrives against a backdrop of eye-watering infrastructure spending. Meta has raised its 2026 capital expenditure forecast to between $125 billion and $145 billion.

The company has pointed to rising component costs and AI capacity needs as reasons for the higher range. In Q2 2026 alone, Meta reported capital expenditures of $31.1 billion.

What this means for Meta and its investors

The core tension is simple. Agentic AI is more useful than a chatbot, but it is also more expensive to run, and those costs scale with every new user.

That is why the ten-times efficiency target is the number to watch. It is a promise about the future cost curve, and it will determine whether Muse becomes a margin story or a margin problem.

The subscription tiers are the second variable. The $20 and $100 price points set up a direct revenue stream, which Meta has rarely relied on in its ad-driven history.

The capex range of $125 billion to $145 billion already leaves little room for surprise, and component costs are part of the reason it went up in the first place.

Li has effectively laid out the bet: costs will rise, efficiency will rise faster, and the company will come out ahead.