Meta Bought Manus AI for $2 Billion in 10 Days. Here's What That Means.

Meta just paid over $2 billion for Manus AI. The deal closed in 10 days. Most acquisitions take months. This one happened so fast that even investors thought it was fake.​

What Manus actually does

Manus is an AI agent that does stuff without you watching it. You give it a task. It goes away and completes it. In the cloud. While you sleep.​

It can screen resumes. Book flights. Build websites from scratch. The difference between Manus and ChatGPT is that ChatGPT waits for your next prompt. Manus just keeps working.​

The company launched in March 2025. By December they had millions of users and were making $125 million a year. That's eight months from zero to nine figures.​

Meta saw that growth and wrote a check for $2 billion before anyone else could.

Why Meta moved fast

Manus was raising money. They wanted a $2 billion valuation. Meta offered them exactly that amount to just buy the whole thing.​

Here's what Meta gets: a working AI agent with real paying customers. Not a demo. Not a research paper. Actual revenue.​

Most AI companies right now are burning cash on compute. Manus was already charging $20 per month and people were paying. Meta gets to plug that into WhatsApp and Facebook immediately.​

And they get the team. Manus founder Xiao Hong becomes a VP at Meta. You don't just buy the code. You buy the people who know how to make it work.​

The China problem nobody wants to talk about

Manus was founded by Chinese entrepreneurs. The company started in China before moving to Singapore.​

Meta says Manus will cut all China ties after the acquisition. No Chinese ownership. No Chinese investors. Clean break.​

People on Reddit are asking if that's even possible. The tech is already built. The founders are who they are. You can move headquarters but you can't rewrite history.​

This is the first time a major US tech company bought a startup with Chinese roots. That alone tells you how badly Meta wanted this.​

The regulatory scrutiny is going to be interesting.

What happens to your Manus subscription

If you're paying for Manus right now, nothing changes immediately. Meta says they'll keep it running as is.​

But let's be honest. They didn't spend $2 billion to keep selling $20 subscriptions. They want this inside WhatsApp. Inside Instagram. Inside those Meta AI glasses nobody bought.​

Give it six months. Maybe a year. Then Manus features start showing up as "Meta AI Pro" or whatever marketing names them.

The standalone product probably sticks around for enterprise customers. Meta loves enterprise revenue. It's steadier than ads.​

Microsoft already tried this

Microsoft partnered with Manus in October. They tested it on Windows 11. Let users build websites from local files.​

Then Meta bought the whole company two months later.

I bet someone at Microsoft is annoyed. You spend time integrating a partner's tech. Then your competitor just buys them. Now you're back to square one.

And that's the game. You either own the AI or someone else will.

Where agents actually matter

Most people don't need an AI agent. They need ChatGPT to rewrite an email. That's it.

But if you run a small business, an agent that books your clients and manages your calendar? That's different. If you're hiring and need to screen 500 resumes? That's real value.​

Manus bet on that use case. They built for people who have repetitive tasks that take hours. Not people who want to chat about philosophy with an AI.

Meta is betting the same thing. They want agents inside WhatsApp Business. Inside Facebook Marketplace. Places where people actually transact.​

The real reason this matters

This is Meta's fifth AI acquisition in 2025. Five. They also dropped $14 billion on Scale AI earlier this year.​

Zuckerberg is not messing around. He's buying every AI company that shows traction. He's hiring researchers from OpenAI and Google.​

Meta spent years being late to mobile. They're not doing that again with AI.

The question on Reddit and HackerNews is whether $2 billion was too much. Manus was valued at $500 million in April. Eight months later Meta paid four times that.​

Some people think it's overpriced. Some think agents are the next platform shift. I think Meta just didn't want OpenAI or Google to get there first.​

The money angle

Meta's stock has been weird about AI spending. Investors keep asking when all this infrastructure pays off.​

Manus gives them an answer. Here's a product. It has users. It makes money. We bought it. Now we scale it.​

That's a cleaner story than "we're spending $50 billion on data centers and trust us it'll work out."

And if agents actually take off, Meta just bought one of the three companies that had real traction. The other two are probably getting acquired next month.

i still think about how fast this deal happened. Ten days. For a $2 billion acquisition. That's not normal. But i guess when you see something working, you just move.