Meta's Acquisition of Manus: A $20 Billion "Execution Capability" Play and New Opportunities in the Computing Power Leasing Market

A bombshell news item has rocked the tech world: Meta has swiftly acquired AI startup Manus for a staggering $20 billion, with the entire deal from initial contact to completion taking just over ten days.

Meta announced the completion of its acquisition of AI startup Manus for up to $20 billion in cash and stock. This marks the third-largest acquisition in Meta's history, trailing only behind its purchases of WhatsApp and Scale AI.

The industry interprets this acquisition as a critical step for Meta to address its "execution capability" shortfall in the AI race. Notably, as Meta channels its massive computing resources towards Manus, the computing power leasing market within the entire AI industry chain may face a new phase of restructuring.

Lightning-Fast Acquisition and Team Integration

Meta's acquisition of Manus was completed with lightning speed, with the entire negotiation process taking just over ten days—a rare pace in the annals of major tech acquisitions. Behind this speed lies the deep usage and high regard that Meta founder Mark Zuckerberg and his executive team have for Manus's product.

According to public information, Manus officially launched its AI Agent product in March 2025, meaning it had only been on the market for about 10 months by the time of the acquisition. In this short period, the company's Annual Recurring Revenue (ARR) had already surpassed $100 million, boasting millions of paying users and establishing itself as one of the fastest-growing star companies in the AI application layer.

The post-acquisition organizational structure is equally noteworthy. Manus's team of approximately 100 people will be fully integrated into Meta's Superintelligence Lab in Singapore. However, founder Xiao Hong will assume the role of Vice President at Meta, reporting directly to Chief Operating Officer Javier Olivan, rather than to Meta's AI chief Yann LeCun (often referred to in media as "Alexander Wang" in some contexts), as many outsiders had speculated.

This unique reporting structure reflects Meta's high regard for the Manus team, allowing it to maintain a relatively independent operational space while ensuring direct communication channels with Meta's core management.

The Core of the Acquisition: Addressing the Execution Capability Gap

From a strategic perspective, Meta's core motivation for this acquisition is to address its "execution capability" gap in the AI field. Despite massive investments in AI infrastructure and model development, Meta has long lacked breakthroughs in user-facing AI applications.

Unlike traditional generative AI chatbots, Manus's product is positioned as a general-purpose AI agent capable of executing multi-step tasks end-to-end. The company's cloud-based virtual machine architecture can create "virtual computers" that open files, run code, and coordinate multiple AI models to accomplish tasks collaboratively.

This capability is precisely what Meta's existing AI products lack. With the integration of Manus, Meta will possess genuine autonomous action capabilities. Users on WhatsApp or Instagram will not only be able to ask, "Help me find travel guides," but can also directly command, "Book my flights, hotels, and generate an itinerary."

By embedding Manus's execution capabilities into WhatsApp, Instagram, and Facebook, Meta is building a new AI ecosystem that can not only chat but also help users "get things done."

Profound Impact on the AI Agent Landscape

Meta's acquisition of Manus has had a profound impact on the entire AI agent landscape. On one hand, this acquisition sets a value benchmark for AI application-layer companies, proving that even without proprietary large language models, AI agent companies with engineering implementation capabilities and clear commercialization paths can become strategic assets coveted by giants.

On the other hand, the acquisition has also raised concerns within the industry about the independent growth space for startups. When giants are willing to acquire leaders in a particular sector at high premiums, the independent growth space for other startups may be further squeezed.

Manus's rapid development path provides a new reference model for AI startups. The company completed the entire cycle of "launch-growth-acquisition" in just 10 months from its official product release to being acquired by a giant—a journey many investors and entrepreneurs describe as a rare "business case study."

This acquisition signifies that AI competition is evolving from a battle at the model level to a confrontation of "ecosystems and resources." Having a smart "brain" (large models) is no longer sufficient; what's needed is a "body" and "operating environment" capable of supporting its effective operation—meaning a complete ecosystem and critical resource support, including computing power.

Computing Resources and New Opportunities in the Leasing Market

Meta's acquisition of Manus could have a profound impact on the computing power leasing market. With the explosive growth in demand for AI Agent applications, computing resources have become a critical bottleneck constraining the development of AI application-layer companies.

Before its acquisition, Manus had created over 80 million virtual computers, processing more than 147 trillion tokens—a feat requiring robust computing power scheduling capabilities. As a startup, Manus continuously faced mounting pressures from escalating computing power and infrastructure costs.

Within Meta's ecosystem, Manus can leverage Meta's powerful Llama model foundation and computing clusters to achieve parallel processing at a much larger scale. This means Manus will no longer need to expend excessive energy on fundraising and procuring basic computing power.

Meta has publicly committed to investing hundreds of billions of dollars in expanding its data centers to support AI-related workloads. These massive computing infrastructure resources will not only serve Meta's own AI applications but may also be opened to external parties in the future through computing power leasing, forming a new business model.

With the integration of Meta and Manus, the computing power leasing market may witness a new competitive landscape. Large tech companies possess the advantage of building their own computing clusters, while small and medium-sized AI companies tend to prefer leasing necessary computing resources. By acquiring Manus to strengthen its AI application capabilities, Meta is also laying the groundwork for potential future computing service offerings.

Future Challenges and Strategic Prospects

Although Meta's acquisition of Manus is seen as a strategic "gap-filling" move, the integration of the two parties still faces numerous challenges. On the technical front, Manus currently primarily uses third-party models such as Anthropic's Claude and Alibaba's Qwen as its underlying capabilities. It remains unclear whether these partnerships will continue post-acquisition or if Meta will push Manus to transition to its own Llama models.

On the market front, Manus's product and team inherently have a toC (business-to-consumer) DNA, with its $125 million ARR revenue structure also primarily toC. However, Meta's focus with this acquisition is expanding services for enterprise clients. How to effectively transform Manus's capabilities into enterprise-grade solutions remains uncertain.

On the competitive front, while Meta holds advantages in infrastructure and computing power, it still lags behind competitors like Google and OpenAI in terms of core model performance, stability of top-tier talent teams, and the coherence of long-term strategy. If the competitiveness of its models cannot catch up, expansion at the application layer will ultimately be constrained.

In the long run, the ultimate success of this acquisition may not depend on whether the deal itself was "worth it," but rather on Meta's ability to effectively synergize model iteration, product integration, and ecosystem collaboration to form a cohesive force. For the broader AI industry, this acquisition marks the opening of a new era, with AI application-layer companies becoming the new focal point of competition among giants.

As Meta deeply integrates Manus's AI agent capabilities into its social media empire, users will experience unprecedented intelligent services on WhatsApp and Instagram. Imagine planning a trip and not just receiving a list of recommendations, but getting confirmed flight tickets, hotel bookings, and a complete itinerary directly.

The ripple effects of this acquisition are spreading. AI startups worldwide are beginning to reassess their value propositions, while players in the computing power leasing market are closely watching whether Meta will open up its vast computing resources. The spring of AI agents may not have fully arrived, but the union of Meta and Manus undoubtedly shines a light on this promising track.