In a major strategic move to accelerate its position in the artificial intelligence race, Meta has announced the acquisition of Singapore-based AI startup Manus. The deal, reportedly valued at over USD 2 billion, represents one of Meta's largest acquisitions to date and signals a decisive pivot towards developing practical, action-oriented AI systems. This acquisition is not merely about acquiring talent or technology; it's about buying a proven path to a future where AI doesn't just answer questions but completes complex tasks autonomously.
The Acquisition and Its Strategic Significance
Meta confirmed the acquisition on December 30, 2025, marking a significant investment in the burgeoning field of AI agents. While the exact financial terms were not officially disclosed, multiple reports, including one from The Wall Street Journal, place the value at over USD 2 billion, with some sources suggesting a figure between USD 20-30 billion. This makes it Meta's third-largest acquisition, trailing only the historic purchases of WhatsApp and Instagram. The deal underscores CEO Mark Zuckerberg's intensified focus on AI as the company's top strategic priority, a shift from its earlier metaverse-centric vision. Following the acquisition, Manus will continue to operate independently, and its co-founder and CEO, Xiao Hong, will join Meta as a Vice President, reporting directly to Chief Operating Officer Javier Olivan.
Acquisition Details:
- Acquirer: Meta (formerly Facebook)
- Target: Manus (Butterfly Effect Technology)
- Reported Value: Over USD 2 billion (estimates range USD 20-30B)
- Rank in Meta's History: Third-largest acquisition (after WhatsApp & Instagram)
- Key Personnel Move: Manus CEO Xiao Hong becomes Meta VP, reporting to COO Javier Olivan.
- Post-Acquisition Structure: Manus continues independent operations.
Manus: From Viral Sensation to Meta's AI Vanguard
Manus, originally founded in China before relocating its headquarters to Singapore in mid-2025, captured the tech world's attention earlier in the year with the launch of what it promoted as the first truly general-purpose AI agent. Unlike conventional chatbots like ChatGPT that provide information, Manus's agent is designed to execute tasks. It operates within a dedicated virtual computer environment, allowing it to interact with software, browse the web, fill out forms, analyze data, and even write and execute code to solve problems. This architecture—a "large model + cloud virtual machine"—effectively gives AI "hands and feet." The company had achieved an annualized revenue run rate of USD 125 million at the beginning of 2025 through a subscription model, providing Meta with an immediate and scalable commercial AI product.
Manus Company Profile:
- Core Product: General-purpose AI Agent that executes tasks in a virtual computer environment.
- Business Model: Subscription-based.
- Annualized Revenue (Early 2025): USD 125 million.
- Headquarters: Singapore (relocated from Beijing, China in June 2025).
- Previous Valuation (Pre-acquisition): ~USD 5 billion.
The "Agent" Revolution: From Information to Action
The core philosophy behind Manus represents a fundamental shift in AI interaction. The industry has recognized that the chatbot interface, while revolutionary, is often not the endpoint for productivity. A chatbot provides an answer, but a user must still manually execute the steps. An agent, however, aims to deliver the final result directly. Xiao Hong illustrated this with a compelling anecdote: when asked to check a train schedule during a website maintenance period, the Manus agent didn't just return an error. It proactively searched for contact information and began drafting an inquiry email, demonstrating goal-oriented persistence. This transition from delivering information to taking action is seen as the next major leap, potentially reshaping workflows and software interaction paradigms.
Meta's AI Roadmap and Competitive Landscape
For Meta, this acquisition addresses a critical gap. Despite massive investments in its open-source Llama large language models and the widespread integration of its Meta AI assistant across Facebook, Instagram, and WhatsApp, the company has lacked a standout, practical AI product for consumers and businesses. The Manus acquisition provides a ready-made, battle-tested agent technology that can be integrated across Meta's ecosystem of apps. Furthermore, by allowing Manus to continue selling its service independently, Meta gains a valuable revenue stream to further fuel its AI ambitions, which include the stated goal of being the first to achieve "superintelligence." This move directly positions Meta against other tech giants like OpenAI, Google, and Anthropic, all of whom are racing to develop their own agentic AI systems.
Technical & Market Context:
- AI Paradigm Shift: Manus represents a move from Information-delivering Chatbots to Action-taking Agents.
- Key Differentiator: "Large language model + Cloud virtual machine" architecture, enabling software interaction.
- Meta's AI Strategy Context: Follows major investments in Llama models, the Meta AI assistant, a USD 14.3B investment in Scale AI (June 2025), and a partnership with Midjourney.
- Competitive Landscape: Places Meta in direct competition with OpenAI's "Operator," Google, and Anthropic in the development of practical AI agents.
Implications for the Future of Work and Software
The technology pioneered by Manus hints at profound changes ahead. If AI can reliably operate software and complete multi-step tasks, the role of the traditional software engineer may evolve into that of an architect or supervisor. Xiao Hong noted that within his company, nearly 80% of code was already AI-generated. This trend could lead to dramatically faster software iteration and pose an existential challenge to traditional SaaS companies that fail to adapt. The future, as some in the industry suggest, may trend toward "There's No Software," where non-standard needs are met not by purchasing new applications, but by instructing an agent to create a solution on the fly. For Meta, acquiring Manus is a bold bet on this very future, aiming to ensure it remains at the forefront of the next decade of AI evolution.