Mark Zuckerberg's Meta Is Charging Consumers for a Personal AI Agent for the First Time. Here's What $20 and $100 a Month Could Add to Revenue.

Meta Platforms (NASDAQ:META) launched Muse on Sept. 8 -- a personal artificial intelligence (AI) agent that can send emails, book travel, fill out forms, and negotiate on your behalf. The app is free, but heavier users can pay $20 or $100 a month for more capacity.

Muse is the company's first personal AI agent, and the paid plans mark the first time Meta has charged consumers for one.

Advertising made up about 98% of Meta's $60.8 billion in second-quarter revenue. How much could $20 and $100 subscriptions add to a machine like that?

Less than you might think. And Meta's own executives are setting expectations low.

Image source: Getty Images.

A new kind of subscription

Muse is built on the company's Muse Spark AI model and is available to U.S. users on iOS and Android, at muse.ai, and through WhatsApp.

Unlike a chatbot, the agent does things for you. It can open a browser on its own, and it keeps working after you close the app.

Meta has charged consumers before, of course. Meta Verified launched at $11.99 to $14.99 a month back in 2023. And this spring the company started testing Meta One, plans at $7.99 and $19.99 a month for heavier use of its Meta AI assistant.

But Muse's top plan costs five times as much as the priciest of those.

That said, most users may never pay anything. AI chief Alexandr Wang has said the vast majority of users should be able to do what they need within the free tier. The paid plans, he added, exist largely to cover computing costs for the heaviest users.

How much revenue are we talking about?

Every 1 million people paying $20 a month works out to about $240 million a year. Every 1 million on the $100 plan adds about $1.2 billion.

Against an advertising business generating about $60 billion a quarter, even 5 million people on the $20 plan would amount to about half of 1% of annual revenue.

Meta's own history shows how slowly these lines build. In the second quarter, the "other revenue" line in Meta's family of apps segment (the line that holds its consumer subscriptions plus WhatsApp paid messaging fees) topped $1 billion in a quarter for the first time, chief financial officer Susan Li said on the July earnings call. That was 73% growth from the year-ago quarter's $583 million, and it followed about $884 million in this year's first quarter. The dollars are stepping up quickly. But after three years of Meta Verified and every other plan the company sells, the line still makes up well under 2% of total revenue.

In other words, the line is Meta's fastest-growing revenue source and still close to invisible in the results. Muse's paid plans will land in that same line, and I don't expect them to transform it.

Meta priced the tiers to cover costs

I think the paid plans look less like a profit engine and more like a way to make the heaviest users pay for the computing they use. And that computing is expensive. After all, Meta expects this year's capital expenditures to come in between $130 billion and $145 billion.

Indeed, the spending is already showing up in Meta's cash flow. Second-quarter capital expenditures alone were $31.1 billion, and free cash flow shrank to $784 million from $8.5 billion a year earlier. A few million $20 subscriptions would barely register against numbers like that.

Why bother with paid tiers, then? Because subscriptions are only the first way agents could make money.

"[W]e're developing new personal agents that will be the foundation for our next wave of products and revenue lines in the months and years ahead," CEO Mark Zuckerberg said on the company's July earnings call.

And Wang told CNBC that Meta is exploring taking a cut of shopping purchases completed through the agent, though it hasn't settled on a plan.

Ultimately, Muse's paid plans won't meaningfully change Meta's revenue anytime soon -- and the company all but said so at launch. The investment case still rests on the ads business, which grew 27% from a year ago last quarter and is helping fund the agent effort.

With shares around $658 as of this writing, the stock sells for about 19 times the earnings per share that analysts forecast for 2027, which is arguably a reasonable price for a business growing this quickly.

So I’d still buy Meta shares here because of the overall value you’re getting relative to the price -- not just because of a $20 subscription.

Daniel Sparks and his clients do not have positions in any of the stocks mentioned. The Motley Fool has positions in and recommends Meta Platforms. The Motley Fool has a disclosure policy.