Skip to main content
Meta AI agent Muse interface on a smartphone, displaying user engagement metrics and AI-generated content.

Editorial illustration for Meta's AI Agent Muse Draws 500,000 Users in First Week

Meta's AI Agent Muse Hits 500K Users in Week

4 min read

Half a million people downloaded an app that admittedly copies its homework, and did it anyway. Meta's personal AI agent Muse launched in the US on September 8 and by the end of its first week had pulled in more than 500,000 users, with over 250,000 opening it daily and collectively firing off more than two million prompts, according to internal data cited by The Information. The app climbed to number one on Apple's App Store, racking up over 31,000 ratings along the way. Muse is built as a standalone app and also runs through WhatsApp, positioned as a stand-in for the tedious parts of daily life: booking trips, hunting deals, tracking expenses, untangling scheduling conflicts.

The numbers look like a hit by any normal startup standard. But the launch came with an unusual admission from Meta itself, one that has users combing through file names and code looking for receipts. The company says Muse draws heavy inspiration from OpenClaw, an open-source project that predates it. How heavy that inspiration turns out to be is where the story gets complicated.

Meta's new AI agent is off to a strong start, but Meta itself says the product is "heavily inspired" by the open-source project OpenClaw. Some of the file names and contents are nearly identical.

Why this matters

Half a million users in a week is a real number, but it tells us more about Meta's distribution than about Muse's staying power. Meta can put an app in front of billions of Instagram and WhatsApp accounts; that's not a signal that agentic AI has found product-market fit, it's a signal that Meta has an audience. The OpenClaw admission is the more interesting data point for developers and founders watching this space.

Nat Friedman crediting the open-source project by name, rather than Meta quietly absorbing the idea, suggests the fastest path to shipping an AI agent right now is copying what's already working in the open-source ecosystem and wrapping it in a bigger company's reach. For researchers and builders, that's worth sitting with: the technical bar for "novel" agent products may be lower than the marketing suggests. The real test isn't install numbers, it's retention, whether people keep asking Muse to book appointments a month from now, and whether Meta's version outpaces the project it borrowed from once the novelty wears off.

Common Questions Answered

How many users did Meta's AI agent Muse acquire during its first week of launch?

Meta's personal AI agent Muse pulled in more than 500,000 users by the end of its first week after launching in the US on September 8. The app achieved significant engagement with over 250,000 daily active users and collectively generated more than two million prompts during this initial period.

What is the relationship between Meta's Muse and the open-source project OpenClaw?

Meta acknowledged that Muse is "heavily inspired" by the open-source project OpenClaw, with some file names and contents being nearly identical between the two projects. This admission raised questions about the extent of copying involved in Muse's development.

What does Meta's rapid user acquisition for Muse indicate about agentic AI adoption?

While 500,000 users in a week is a significant number, it primarily demonstrates Meta's powerful distribution capabilities rather than proof that agentic AI has achieved product-market fit. Meta's ability to promote the app through its billions of Instagram and WhatsApp users accounts for the rapid adoption more than the product's inherent appeal.

What was Muse's ranking on Apple's App Store following its launch?

Muse climbed to number one on Apple's App Store shortly after its launch, accumulating over 31,000 ratings in the process. This top ranking reflects both the app's initial popularity and Meta's significant promotional reach.

LIVE19:46Google's Flash TTS Lets Users Design AI Voices From Scratch