Meta's New AI App Muse Hits #2 in the US — So Why Is Growth Still Sluggish?
Meta's new AI agent app Muse ranks #2 in the US but grows slower than expected.
Meta launched its AI agent app Muse, which quickly reached #2 on the US App Store — but early growth has been noticeably slower than products like Threads. This reflects insufficient ecosystem-driven traffic, high user education costs, and waning novelty. In a fiercely competitive AI agent market, Muse's long-term value will depend on user retention and product quality, not short-term download rankings.
Meta Launches a New AI Agent App: Muse
Meta recently quietly launched a new AI agent app called Muse, which quickly climbed to #2 on the US App Store download charts. On the surface, that sounds impressive — but a closer look at its growth curve reveals that Muse is actually off to a slower start than many of Meta's other flagship products.
As a tech giant with simultaneous stakes in social media and AI, Meta has been moving fast. From Meta AI — which integrates large language model capabilities — to Threads, its Twitter rival, Meta has poured enormous resources and traffic support into nearly every new product line. Muse, as the latest move in Meta's AI agent strategy, carries the company's ambitions to expand further into the generative AI application layer.
What the "#2 Ranking" Actually Tells Us About Muse's Growth
Despite Muse currently sitting at #2 on US app charts, that ranking doesn't fully capture how well users are actually embracing it. According to reports, Muse's early growth has been noticeably slower compared to previous Meta launches like Meta AI or Threads.
Comparing Against Threads' Explosive Start
Look back at Threads' launch in 2023: leveraging Instagram's massive user base, it surpassed 100 million registered users in just five days — a record for consumer app growth. That playbook of "riding the social ecosystem" is Meta's go-to growth engine, and it's been remarkably effective.
Muse hasn't replicated anything close to that explosive start. This could suggest several things:
- Insufficient ecosystem-driven traffic: Muse hasn't received the same level of social ecosystem funnel support that Threads did;
- Higher user education costs: AI agent products have a steeper conceptual learning curve, and retention after the initial trial period remains uncertain;
- Fading novelty: Users are growing less excited about "yet another AI app."
The Disconnect Between App Store Rankings and Real Growth
It's worth noting that App Store download rankings primarily reflect short-term download momentum — not long-term engagement. Even if an app shoots to the top of the charts, its commercial value remains limited without sustained user retention and depth of usage. Muse's current situation — high ranking, slow growth — exposes the gap between raw download numbers and genuine user stickiness.
The AI Agent Race: Competitive Landscape and Core Challenges
AI agents are widely seen as the next major battleground in generative AI, following the chatbot wave. Unlike traditional conversational AI, agents emphasize the ability to proactively execute tasks — not just answering questions, but carrying out sequences of actions on the user's behalf.
A Crowded, Competitive Market
Right now, OpenAI, Google, Anthropic, and a wave of startups are all competing for dominance in the AI agent space. Muse is entering a market that's already quite crowded. With products like ChatGPT and Gemini already having established strong user mindshare, breaking through won't be easy for Muse.
Meta's unique advantage lies in its social platforms' massive user base and contextual data. In theory, Muse could deeply integrate with Instagram, WhatsApp, Messenger, and other ecosystems to carve out a differentiated competitive position. But given the sluggish growth so far, Meta either hasn't fully unleashed those synergies yet — or is deliberately taking a more cautious rollout approach.
User Education Remains a Long-Term Challenge
One of the core challenges for AI agent products is getting everyday users to understand and get comfortable with the entirely new usage pattern of "delegating tasks to AI." This is fundamentally different from the search and chat experiences people are already used to. Muse's slow start, to some extent, also reflects the fact that the entire industry is still in early stages when it comes to building new user habits.
A Broader View of Meta's AI Strategy
Zooming out, Muse is just one piece of Meta's much larger AI strategy. Meta's sustained investment in its open-source Llama model series has already made it a significant player at the AI infrastructure layer. On the application layer, Meta is continuously probing the boundaries of the consumer market through products like Meta AI and Muse.
This dual-engine strategy — infrastructure plus applications — lets Meta explore opportunities simultaneously at different levels. Even if a product like Muse gets off to a slower-than-expected start, Meta has the patience and resources to keep iterating. After all, for a company of Meta's scale, short-term growth numbers for a single product matter — but what matters more is whether it can stake out a position in the AI agent space that will define the next era.
Closing Thoughts: Beyond the Rankings, Look at Long-Term Value
Muse hitting #2 in the US is an achievement worth acknowledging — but the "slow start" signal is equally impossible to ignore. For an emerging category like AI agents, the real test isn't first-week download numbers. It's whether users keep coming back and whether the product builds irreplaceable value in their lives.
For Meta, Muse is an important market-probing exercise. Whatever the final outcome, it will provide valuable data feedback on what consumers actually want from an AI agent. And for the broader industry, Muse's performance serves as yet another reminder: in the AI application layer, the era of easy traffic gains is fading. Product quality and genuine user value are what ultimately determine success.
Related articles
LangChain + MCP: From Core Concepts to Agent Tool Calling in Practice
Learn how LangChain and MCP work together — covering LLM tool calling, Agent architecture, and conversation history management to build real-world AI applications.
Probabilistic Machine Learning: Why It's the Cornerstone to Unlocking the ML Black Box
Without probability theory, ML is always a black box. This article explores why probabilistic foundations are essential for understanding machine learning algorithms, Bayes' theorem, MLE, and more.
Optimization Pitfalls in Self-Evolving LLM Agents: Value Concentration and Budget-Splitting Problems
HARNESSEVO research reveals 3 key LLM agent harness optimization findings: value concentrates in reflection/control slots, uniform budget splitting is harmful, and credit assignment must precede structured evolution.