Meta's Muse AI Agent Tops US iOS App Charts as Company Seeks to Rebuild Consumer Trust
Key Takeaways
- •Meta's Muse AI agent launched on September 8 and quickly reached the number-one position on the US iOS free app charts.
- •Unlike conventional chatbots, Muse performs actions such as sending emails, booking travel, and completing purchases through integrations with Gmail, Spotify, and Stripe.
- •Muse is offered in a free tier, a $20-per-month Power plan, and a $100-per-month Maximum tier priced the same as OpenAI's ChatGPT Pro.
- •Meta's stock has risen approximately 27% since the launch, with some analyst estimates placing the gain closer to 36%.
- •Each Muse session runs inside a sandboxed virtual machine, and the agent seeks explicit user permission before handling payment details or personal data.
Meta has accomplished something unusual among Big Tech firms: launching an artificial intelligence product that people genuinely want to use. Muse, the company's personal AI agent, went live on September 8 and promptly claimed the top spot on the US iOS free app charts.
The timing is notable. Muse's debut follows Meta's $18 billion multistate settlement over social media harms to young users, and the product has been introduced with a privacy-first message. That framing carries practical weight: an agent built to reach into email inboxes and payment flows depends on user trust in a way that a text-based chatbot does not.
What Muse Actually Does
Where the wave of chatbots that followed ChatGPT's arrival mostly generate text, Muse is built to take action. The agent sends emails, books travel, and handles purchases, connecting to services that include Gmail, Spotify, and Stripe. That shift from generating text to executing tasks is where much of the industry's competitive energy is now directed—and it is also what raises the stakes on reliability and security.
Under the hood, the system runs on what Meta describes as Muse Spark models. Every agent session takes place inside a Muse Secure VM, effectively a sandboxed virtual machine that provides the AI with browser access for tasks lacking a direct API. Whenever the agent needs to carry out a sensitive operation—entering payment details or reaching personal data—it first asks the user for explicit permission.
The pricing follows the tiered playbook now common in the AI industry. A free version covers basic tasks, a $20 per month Power tier unlocks expanded capabilities, and a $100 per month Maximum subscription delivers the full suite. That top tier matches what OpenAI charges for ChatGPT Pro.
Wall Street's Verdict
Meta's shares have climbed roughly 27% since Muse launched, and some analyst estimates put the gain closer to 36%. The subscription structure also hands analysts something concrete to model. Meta has historically been a pure advertising business, and recurring revenue from Muse subscriptions diversifies that exposure in a way investors have been asking for. Whether the chart-topping launch converts into durable subscription revenue will become clearer in Meta's quarterly disclosures.
Meta Connect 2026 added fuel to the rally by previewing further Muse features, among them digital avatars and expanded service integrations.
The Competitive Landscape Shifts
Meta's AI infrastructure buildout, led by Chief AI Officer Alexandr Wang through Meta Superintelligence Labs, gives the company a technical foundation that few rivals beyond Google and Microsoft can match. The compute demands are concrete: every Muse session spins up its own sandboxed virtual machine, a heavier per-user commitment than serving text responses alone.
Notably, Muse arrived as a standalone app rather than being embedded in Meta's existing products. It therefore has to win downloads on its own merits instead of leaning on pre-installed distribution across Facebook, Instagram, and WhatsApp. So far, it is doing exactly that.
