Meta Acquires Manus as AI Agents Move Into Business Workflows

When Meta confirmed it was acquiring Singapore-based AI startup Manus, the announcement landed less as a surprise and more as a signal. After years of pouring billions into artificial intelligence infrastructure, Meta appears to be shifting its focus from building intelligence to operationalising it.

Manus does not build another large language model. Instead, it positions itself as an execution layer, an autonomous AI agent designed to carry out end-to-end work across research, automation, and complex business tasks. For Meta, that distinction matters.

As pressure mounts to justify the scale of its AI spend, the Manus deal offers something increasingly scarce in the AI market: a product that already sells, already scales, and already fits inside business workflows.

Why Manus Fits Meta’s AI Pivot

In a statement, Meta said it would “operate and sell the Manus service” while integrating its technology into Meta products such as the Meta AI chatbot. Financial terms were not disclosed.

Manus sells its AI agent through a subscription model, with businesses able to purchase access for as little as $20 a month. Earlier this year, the company had an annual revenue run rate of $125 million, an important data point at a time when many AI platforms are still struggling to translate usage into revenue.

Manus released its AI agent earlier this year, capable of handling a range of general tasks, including screening resumes, creating trip itineraries, and analysing stocks from basic instructions. The focus is not novelty but repeatability, automating work that businesses already do.

This aligns with Meta Chief Executive Officer Mark Zuckerberg’s broader push to build a sustainable AI business. While Meta continues to compete with OpenAI and Google on foundational models, agents offer a faster route to monetisation.

Inside Manus: Execution Over Experimentation

In its own blog post titled “Manus Joins Meta for Next Era of Innovation”, the company framed the acquisition as validation of its core thesis.

https://manus.im/blog/manus-joins-meta-for-next-era-of-innovation

“Since the launch, Manus has focused on building a general-purpose AI agent designed to help users tackle research, automation, and complex tasks,” the company said, noting that the agent has already processed more than 147 trillion tokens and powered the creation of over 80 million virtual computers.

The emphasis, however, was on reliability rather than scale for scale’s sake.

“We believe in the potential of autonomous agents, and this development reinforces Manus’s role as an execution layer, turning advanced AI capabilities into scalable, reliable systems that can carry out end-to-end work in real-world settings,” the blog post said.

Crucially for existing customers, Manus said it would continue to operate independently from Singapore and keep selling subscriptions through its app and website.

“Our top priority is ensuring that this change won’t be disruptive for our customers,” the company said.

A Cross-Border Deal With Global Implications

Manus’s journey adds another layer of complexity to the acquisition. Founded in China before relocating to Singapore, the company’s parent, Butterfly Effect, raised capital earlier this year at close to a $500 million valuation, with Benchmark among its investors.

The deal comes at a time when AI has become a flashpoint in US-China relations. Investments involving Chinese-founded AI teams are increasingly scrutinised in Washington, while talent migration continues to spark debate in Beijing.

An early investor in Manus said the acquisition would rank as Meta’s third-largest deal after WhatsApp and Scale AI. Meta paid $19 billion for WhatsApp in 2014 and up to $15 billion for a 49% stake in Scale AI earlier this year.

“The era belonging to this generation of young Chinese entrepreneurs has arrived,” said Liu Yuan, Partner, ZhenFund, noting that Zuckerberg has been a long-time Manus user.

Others were more cautious. Li Chengdong, founder of the Haitun Internet Think-Tank, said, “An incredible company and team are being sold to the United States,” warning that failure to value entrepreneurial talent could hurt China’s position in the global AI race.

AI Agents as Meta’s Revenue Bridge

Beyond geopolitics, the Manus acquisition reflects a broader industry shift. As AI infrastructure costs rise, companies are under pressure to show near-term returns. Autonomous agents—unlike general-purpose chatbots—are easier to price, package, and justify to enterprises.

Meta has already been exploring premium subscriptions for its AI assistant, Meta AI, including features such as booking reservations and video creation. Manus provides both a working product and a commercial playbook.

AI entrepreneur Alexandr Wang, recently hired by Zuckerberg to lead Meta’s AI efforts, said Meta is recruiting more widely in Singapore, where the Manus team will remain based.

For Zuckerberg, who has spoken openly about building “personal superintelligence”, the path forward increasingly runs through practical deployments rather than abstract capability gains.

The Manus acquisition underscores a quiet recalibration in the AI market. The next phase of competition may not be won by who trains the biggest model, but by who turns AI into dependable, revenue-generating systems.

“Joining Meta allows us to build on a stronger, more sustainable foundation without changing how Manus works or how decisions are made,” said Xiao Hong, Chief Executive Officer, Manus. “We’re excited about what the future holds with Meta and Manus working together, and we will continue to iterate the product and serve users that have defined Manus from the beginning.”

For Meta, the message is equally clear: intelligence alone is no longer enough. Execution and monetisation are now the real battleground.