Meta has pulled its Muse AI assistant out of the Facebook-and-Instagram scaffolding and given it a home of its own. Muse is now a standalone product, reachable through a dedicated app, the web, and WhatsApp, with support for Meta’s AI glasses on the way. There is a free tier, a $20-a-month plan and a $100-a-month plan, which is the now-standard three-rung pricing ladder every assistant seems to arrive at.
The move itself is the story. For two years Meta’s AI lived inside its other apps, a helpful ghost in the machine you did not go to on purpose. Cutting it loose as its own destination is Meta admitting that the assistant has to compete head-on with ChatGPT, Gemini and Claude as a place people choose to open, not just a feature they stumble into while scrolling. That is a much harder game, and a more honest one.
Why standalone matters
Distribution is the one thing Meta has that OpenAI would arguably kill for: billions of people already inside its apps every day. Baking Muse into that flow guaranteed reach but guaranteed nothing about intent, because a feature you meet by accident is not a habit. A standalone app, plus a web presence and WhatsApp access, is a bet that Muse is good enough to earn a spot on your home screen next to the incumbents, rather than relying on being unavoidable.
The glasses angle is the part worth watching. A voice assistant on your face, hands-free and always a sentence away, is a genuinely different proposition from a chat box, and it is the one form factor where Meta has a real hardware lead over its software rivals. If Muse becomes the default brain behind the glasses, the standalone app starts to look less like a me-too ChatGPT clone and more like the phone-shaped front door to something Meta actually owns end to end.
For now, the honest read is that this is a distribution and pricing move rather than a leap in capability. Whether Muse is worth opening on purpose comes down to how it performs against assistants people already trust, and that is a question the free tier lets you answer yourself. Our advice: try it on the glasses use-case, where it is differentiated, before judging it as yet another chatbot, where it is not.
Related on Top Tool Stack: Meta’s internal Hatch agent.