Meta Muse: What Meta's New Personal AI Agent Actually Does

Meta launched Muse, its personal AI agent, to US users on September 8, 2026. Developed internally under the code name Hatch, Muse runs on Meta's Muse Spark family of models and is built to take everyday digital tasks off your plate.

Here is what Muse does, how its privacy model works, what it costs, and where it is available, based on Meta's own announcement.

What Muse can do

Muse is an agent, not just a chatbot. According to Meta, it can:

• Book appointments
• Fill out forms
• Monitor home security cameras

Meta positions it for the routine digital chores people tend to put off, and it carries out those tasks on your behalf rather than only answering questions.

The privacy model: an isolated environment

Meta's pitch for Muse leans on safety. Meta says Muse runs in an isolated, secure environment and has no visibility into your passwords or payment methods. Before it takes a sensitive action, such as sending an email or making a purchase, it checks with you first.

Pricing

Muse is free for most everyday use. Two paid subscriptions are available for heavier use:

• Free tier
• $20 per month
• $100 per month

Meta says pricing scales with usage.

Availability

At launch, Muse is limited to US consumers aged 18 and over. It is available on iOS, Android and a standalone site, and Meta says more countries will be added over time.

Muse Code is a different product

Do not confuse Meta Muse with Muse Code. Muse Code is a separate Meta developer CLI aimed at engineers. Meta Muse is the consumer personal-agent app covered here.

Bottom line

Meta Muse is a consumer agent for people who want to hand off scheduling, forms and other digital errands, with a privacy model built around an isolated environment and explicit confirmation before sensitive actions. It is US-only at launch, free to start, and priced by usage above the free tier.

See the Meta Muse listing on Toolradar for features, pricing and alternatives.