Meta's AI product Muse hits 700K DAU in 11 days, faces compute bottleneck
Editorial responsibility
- No named human review is recorded for this page.
- Source reporting is collected, normalized, translated or condensed automatically when needed.
- Automatically published source-backed update
Meta's artificial intelligence product, Muse, has reached approximately 700,000 daily active users within 11 days of launch, a tenfold increase, according to Similarweb. However, the rapid growth has exposed severe compute infrastructure limitations. Monitoring platform SaaSHub has listed Muse as experiencing 'service degradation,' with users reporting frequent search failures. A researcher's stress test showed that when asked to generate 120 sub-agents, Muse only successfully created 33. Meta's Chief AI Officer Alexander Wang stated on September 9 that early usage far exceeded internal expectations, roughly ten times the beta test level, prompting Meta to reset token usage limits. The bottleneck is attributed to Muse's compute-intensive architecture, which allocates each user a dedicated virtual machine with 2 virtual CPUs, 8GB RAM, and 100GB SSD. Industry analysis suggests that if Muse reaches 100 million users within a year without resource sharing, Meta would need approximately 2 billion CPU cores, 800PB of RAM, and 10,000PB of SSD storage. The service is currently only available in the US and Canada, with no WhatsApp or Instagram user registration open yet. Pricing starts at a free tier of up to 100 million tokens per week, with paid subscriptions from $20 to $100 per month.
Source report
September 25 — Meta's artificial intelligence platform, Muse, has experienced explosive user growth but is now grappling with significant infrastructure constraints, according to recent data and reports.
Surge in Daily Active Users
According to statistics from Similarweb, Muse's daily active users surged tenfold over the past 11 days, reaching approximately 700,000. While this growth rate is remarkable, the overall user base remains relatively small within the broader AI application landscape.
Signs of Service Degradation
Despite its modest scale, Muse has already shown signs of strain under the growing demand:
- SaaSHub, a status monitoring platform, has classified Muse's service status as "degraded," noting that while its monitoring system can still access the platform, it consistently reports issues — including frequent failures to return search results.
- A researcher conducted a stress test, requesting Muse to generate 120 sub-agents at once. The platform successfully created only 33, with all remaining attempts ending in failure.
Scalability Concerns
These scalability issues are particularly concerning given that Muse's user base has yet to reach 1 million. Currently, Meta has only launched Muse in the United States and Canada, and has not opened registration channels for WhatsApp or Instagram users. If Meta aims to capture further market share, its current computing power bottleneck will become the most significant constraint.
Internal Expectations Exceeded
Meta's Chief AI Officer, Alexander Wang, revealed on September 9 that early usage of Muse had far exceeded internal expectations — approximately ten times the usage level seen during internal testing. This prompted Meta to reset token usage limits and impose additional restrictions.
Computing-Intensive Architecture
A key factor behind the bottleneck may be Muse's computing-intensive architecture. Meta has committed to providing each Muse user with a dedicated virtual machine, equipped with:
- 2 virtual CPUs
- 8GB of memory
- 100GB of solid-state drive (SSD) storage
Projected Hardware Demands
Industry analysis has previously estimated that if Muse's user base reaches 100 million within a year — with no resource sharing among users — Meta would require:
- At least 200 million CPU cores, equivalent to approximately 1.58 million 126-core AMD Ryzen EPYC CPUs
- 800 petabytes (PB) of memory
- 10,000 PB of SSD storage
Reaching 100 million users is not uncommon for top-ranking AI applications in app stores. The real challenge lies in whether Meta can secure sufficient hardware support.
Pricing vs. Infrastructure Costs
Furthermore, the scale of hardware investment may not align with Muse's current pricing model:
- Free tier: Up to 100 million tokens per user per week
- Paid subscriptions: Starting at approximately $20 per month, up to $100 per month
(Source: Cailianshe)
Source
东方财富网-公司资讯Eastern