According to engadget, Meta has acquired the AI startup Manus in a deal valued at more than $2 billion. The company, originally called Butterfly Effect, launched in March 2025 and quickly claimed to be serving millions of users and businesses. It reportedly hit an annualized revenue run rate of over $100 million in just eight months. Manus is known for building AI agents that automate complex tasks like market research, coding, sales analysis, and website creation. The startup had recently moved its headquarters from Beijing to Singapore and laid off most of its China-based employees. Manus CEO Xiao Hong stated that joining Meta provides a “stronger, more sustainable foundation” for the company.
Meta bets big on agents
So, Meta just dropped a colossal sum on what is essentially a very sophisticated task-automation layer. Here’s the thing: this isn’t just buying a cool model. Manus is an orchestrator. It doesn’t build its own core AI models; it uses others, like Anthropic’s Claude and Alibaba’s Qwen, and strings them together to actually do things. That’s the holy grail right now—moving from a chatbot that gives you instructions to an agent that executes the job. For Meta, which is desperate to prove its AI is useful beyond its own apps, this is a huge shortcut. They’re buying a proven system and, more importantly, a team that knows how to make AI work autonomously. It’s a talent and tech acquisition rolled into one.
Explosive growth and skepticism
But let’s talk about that growth. Over $100 million in annualized revenue eight months after launch? That’s insane. It also raises eyebrows. The engadget piece even includes a skeptic’s quote calling Manus “a product devilishly optimized for influencers.” That’s a fascinating and probably accurate critique. Its initial virality and the types of tasks it automates—website cloning, quick market research—absolutely play into the hands of content creators and hustlers looking for an edge. The question is, is that a sustainable, enterprise-grade foundation? Or is it a flashy demo that struggles with real, messy business problems? Meta’s betting it’s the former, but integrating this “influencer-optimized” tech into large-scale business products could be… tricky.
The China connection and global ambitions
This is also one of the largest exits yet for a startup from China’s AI ecosystem. Manus’s quick pivot to Singapore and the layoffs in Beijing this summer weren’t coincidental. They were cleaning house for a global sale. For a US tech giant like Meta, acquiring a company with deep roots in China’s competitive AI scene is a double-edged sword. They get incredible technical talent and insight, but they also inherit all the geopolitical scrutiny. Basically, Meta isn’t just buying software; it’s buying a bridge—or maybe a lightning rod—between two fiercely competitive AI superpowers. It’s a bold move, for sure.
What happens next
Now, the big test begins. The CEO says nothing will change in how Manus works or makes decisions. I’m skeptical. That never lasts. The real challenge is scale and integration. Can Manus’s agents, which dazzled early adopters, hold up under the load and complexity of Meta’s billions of users? And how does this fit with Meta’s other AI projects, like its Llama models? If they can successfully embed these automation agents into WhatsApp for business, into Instagram shops, or for advertisers, the $2 billion price tag might look cheap. But if it becomes another siloed experiment that loses its magic inside a corporate giant, it’ll be a very expensive lesson. Only time will tell if this is genius or desperation.
