Meta Muse App: Why 10M People Downloaded It in 40 Days

Meta Muse App: Why 10M People Downloaded It in 40 Days

Ten million downloads in forty days. That's how fast Meta's new Meta Muse app passed 10 million users — a pace that puts it in the company of Threads and a handful of other record-breaking launches. You've probably seen it climbing the App Store charts or spotted the ads across Instagram and Facebook. But what actually is this thing, and is it worth a slot on your home screen? We dug through the launch numbers, what it can and can't do, and a privacy incident Meta would probably rather you didn't know about.

Quick answer: The Meta Muse app is Meta's free personal AI agent app (iOS and Android), launched September 8, 2026. Unlike a plain chatbot, it can do things for you — send emails, compare prices while shopping, manage your schedule. It passed 10 million US downloads in 40 days with stronger user retention than ChatGPT, according to Sensor Tower data reported by TechCrunch. It's worth trying if you want an AI that acts, not just answers — but know about the paused chat-sharing feature before you share anything sensitive.

Table of Contents

What is the Meta Muse app?

Muse is Meta's standalone AI assistant app — a separate download on iPhone and Android, free, that works as a personal AI agent rather than a simple question-answering chatbot. The difference matters: where classic chatbots like the early ChatGPT mostly answered questions, an agent is built to do things on your behalf. Tell Muse to find you a flight, compare prices across stores, or draft an email, and it takes steps in secure cloud environments to get it done, asking for your approval before anything sensitive like a payment.

This is the same direction the whole industry is moving. Agentic AI is the big 2026 theme — Notion rebuilt itself around AI agents this year, and every major lab is racing to turn its chatbot into a doer. Meta just happened to package the idea into the most viral launch of the season. If you want to understand the technology family Muse belongs to, our explainer on Meta's open Muse Spark models covers the model-side side of the story.

Why did Meta launch it as a separate app?

Distribution. Meta already has around 3.6 billion daily users across Facebook, Instagram, and WhatsApp, and it used that firehose to promote Muse relentlessly. Sensor Tower estimates Meta directed 39% of its entire US advertising spend to Muse in October 2026, up from 23% in September. When you own the billboards, the app store, and the social feed, 10 million downloads start looking very achievable.

The numbers behind the 10-million milestone

The headline figures come from app-intelligence firm Sensor Tower, first reported by TechCrunch on October 20:

  • 10.9 million total downloads within 40 days of the September 8 launch — 10.2 million of them in the United States alone.
  • One of only about a dozen apps in history to reach 10 million US downloads that fast (Threads, Meta's last record-breaker, was the previous one).
  • 2.5 million daily active users on October 16, growing 15% week over week — so people aren't just installing it and forgetting it.
  • 30% of US users were still active 7 days after install, 23% after 14 days — versus roughly 19% and 16% for ChatGPT, per Sensor Tower's retention data. Retention is the number analysts watch, and Muse is beating the incumbent.
  • 4.8-star rating across 420,000+ ratings on the app stores, and a #1 chart position in free apps in the US, Japan, and Canada (AppInsider chart data).

Meta confirmed the milestone on its third-quarter earnings call, saying the app had passed 10 million users and that people were leaning on it for two things in particular: shopping and personal productivity. For context, Meta's Q3 revenue was $64.6 billion, up 26% year over year, with AI capital expenditure guidance of $130–145 billion for 2026. This company is all-in on AI, and Muse is its consumer beachhead.

What can you actually do with it?

Based on Meta's description and early coverage, Muse handles three buckets of tasks:

1. Shopping

This is the use case Meta keeps highlighting. Muse can research products, compare prices across retailers, and streamline the purchase — with enhanced features for small businesses added in late September 2026. It runs in a secure cloud environment and asks your permission before taking sensitive actions like payments, so it's not one-click buying without you in the loop.

2. Personal productivity

Scheduling help, email drafting and sending on your behalf, travel planning — the classic personal-assistant jobs. If you've ever used AI to plan a trip, Muse is Meta's bet that people want an agent that books and organizes, not just suggests.

3. Everyday questions and creation

Writing help, explanations, brainstorming, voice interaction — the standard assistant toolkit, wrapped in Meta's design and connected to the Meta ecosystem.

Muse vs ChatGPT vs Gemini vs Claude

FeatureMeta MuseChatGPTGemini
PriceFree; paid tiers ($20/$100 per month) reportedly introducedFree tier; Plus $20/moFree tier; Advanced plans
Agent capabilitiesBuilt for tasks: email, booking, shopping, schedulingAgent mode growing; strong on reasoningDeep Google Workspace integration
EcosystemFacebook, Instagram, WhatsApp — 3.6B daily usersStandalone; biggest brand recognitionAndroid, Gmail, Google services
User retention (14-day)23% (Sensor Tower)~16% (Sensor Tower)Not disclosed
Privacy track recordChat-sharing feature paused after exposure incident (see below)Recently added chat locking (Face ID/PIN)Google account controls

Honest take: on raw capability, the frontier models are close enough that your choice comes down to which ecosystem you live in and which agent you trust with your tasks. Muse's edge is Meta's distribution and its agent-first design; its weakness is exactly what the next section is about. For more on the models behind these assistants, see our breakdown of GPT-6's intelligent UI and the Mercury 3 diffusion model.

How to get started in 5 minutes

  1. Download the app. Search "Muse from Meta" on the App Store or Google Play — it's free. It's been topping the free charts in the US, Japan, and Canada, so it's easy to find.
  2. Sign in with your Meta account. The app ties into Meta's ecosystem; your Facebook or Instagram identity carries over.
  3. Try one real task first. Don't just chat — that's what separates this app from a chatbot. Ask it to compare prices on something you're actually shopping for, or to plan next week's schedule.
  4. Review permissions before delegating. Muse asks approval for sensitive actions like payments. Read what it's asking, especially the first time — this is the one habit that keeps agentic AI safe.
  5. Don't use the share-chat feature right now. It's paused anyway, but make it a habit: never put bank details, loan applications, or internal company documents into a sharable AI chat.

The privacy incident you should know about

A week is a long time in AI. Days after celebrating the 10-million milestone, Meta had to pause Muse's conversation-sharing feature after security firm Claroc found roughly 6,800 shared Muse chat links discoverable through public DNS records — no hacking required. Here's what happened:

  • When you shared a Muse conversation, the app generated a web address that embedded the chat's title in the URL.
  • Visiting that address triggered a DNS lookup, and services that record DNS activity kept those lookups — titles and all — in searchable public records.
  • Claroc's passive-DNS search surfaced thousands of chats "like a table of contents," including a small-business loan application, a bank fraud alert with partial account details, a company vacation calendar, and meeting notes involving a leading AI lab.

Meta confirmed the pause, telling The Verge it was reviewing the findings and reconsidering how share links get discovered. Claroc later confirmed the exposed links were closed off. No evidence suggests private, unshared chats were exposed — the leak only affected conversations users had chosen to share. Still, it's a useful rule for every AI app: assume anything you "share" can become public, and never paste financial or corporate secrets into an AI chat. The timing stung for Meta: OpenAI rolled out chat-locking (Face ID, Touch ID, PIN) for ChatGPT in the very same week.

Honest pros and cons

Pros

  • Actually agentic — it completes tasks (shopping, scheduling, email), not just answers questions.
  • Free, with top-chart polish — 4.8 stars across 420k+ ratings says the experience is solid.
  • Best-in-class retention — 30% day-7 retention beats ChatGPT's 19%, suggesting people genuinely keep using it.
  • Meta ecosystem advantage — if your life runs on Instagram, Facebook, and WhatsApp, this slots in naturally.

Cons

  • The sharing incident — a real, documented privacy design failure in its first 40 days. Trust has to be re-earned.
  • US-centric for now — 10.2 of 10.9 million downloads are American; international users are early.
  • It's Meta — the company with the weakest privacy reputation among the AI giants is asking you to let its agent handle your shopping and email. That's a genuine tradeoff.
  • Paid tiers loom — $20 and $100/month subscription tiers are reportedly on the way, so "free" may stratify.

Who should try it — and who should skip it

Try it if: you shop online a lot and want price comparisons done for you; you like the idea of an AI that schedules and sends rather than just suggests; you're deep in Meta's apps already.

Skip it (for now) if: you handle sensitive business or financial data daily and want maximum privacy controls; you're outside the US and want a fully localized experience; you already have an agent workflow in ChatGPT or Gemini you're happy with. There's no wrong answer here — these apps are free to test, and switching costs are basically zero.

Frequently asked questions

Is the Meta Muse app free?

Yes — the app is free to download and use on iPhone and Android. Reports indicate Meta is introducing paid subscription tiers at $20 and $100 per month, similar to ChatGPT Plus-style pricing, so some advanced features may eventually sit behind a paywall.

How is Muse different from ChatGPT?

ChatGPT started as a chatbot that answers questions; Muse was built from day one as a personal AI agent that performs tasks — sending emails, comparing prices, managing schedules — with your approval for sensitive actions. In practice the gap is narrowing as every assistant adds agent features.

Is the Meta Muse app available outside the US?

It's downloadable internationally and has topped free-app charts in Japan and Canada, but 10.2 of its 10.9 million first-40-day downloads were American. International expansion is still early, so some features may feel US-focused.

Is the Meta Muse app safe to use?

For everyday use, yes — but learn from the October 2026 incident: Muse's chat-sharing feature was paused after researchers found about 6,800 shared conversations discoverable via public DNS records. Never share AI chats containing financial details, loan applications, or company documents, in any app.

Can Muse book flights and buy things for me?

It can research, compare, and arrange things like travel and shopping, but Meta says it runs in secure cloud environments and requires your approval before sensitive actions such as payments. Think of it as a very fast assistant that still asks before spending your money.

Do I need a Facebook or Instagram account to use Muse?

The app is tied to Meta's ecosystem, so a Meta account is the way in. That's also why adoption exploded: Meta promoted Muse across Facebook, Instagram, and WhatsApp to its roughly 3.6 billion daily users.

Verdict

The Meta Muse app earned its 10 million downloads honestly — it's a genuinely useful agent with the best early retention numbers in the business. But "most downloaded" and "most trustworthy" are different races, and the DNS-sharing fiasco in week six is a reminder that speed and caution don't always travel together. Install it, try one real shopping or scheduling task, and judge it by whether it saves you time — just keep your bank statements out of the share button. That's the review in one line.

📚 Read next on The AI Tools

About The AI Tools: we cover AI tools, apps, and models with hands-on explainers. Learn more on our About page or contact us.

Last updated: October 24, 2026
Methodology: This article was researched from official reporting and app-intelligence data published October 20–24, 2026, including Sensor Tower figures via TechCrunch, Meta's Q3 2026 earnings remarks, AppInsider chart data, and security-firm Claroc's findings as reported by TechSpot and Dataconomy. Download and retention figures are third-party estimates, not Meta's own disclosures. No sponsorships — The AI Tools is reader-supported and independent.

Comments

Post a Comment