---
title: "Meta 收购 Manus 的根本原因是什么？为什么网上的质疑声这么大？"
type: "Topics"
locale: "en"
url: "https://longbridge.com/en/topics/100000000372028.md"
description: "这事吧，大多数人第一眼就看错了。大家都在说：Manus 是 Agent、是生态位、是套壳。但说实话，Meta 根本不是冲这些去的。你站在扎克伯格的位置想一秒就明白了：他现在最焦虑的根本不是 “有没有一个好看的 Agent 产品”，而是——AI 战争，Meta 在数据和实战经验上落后了。Manus 真正值钱的就两样东西。第一是 工程环境。不是 PPT 里的 demo..."
datetime: "2025-12-30T04:05:57.000Z"
locales:
  - [en](https://longbridge.com/en/topics/100000000372028.md)
  - [zh-CN](https://longbridge.com/zh-CN/topics/100000000372028.md)
  - [zh-HK](https://longbridge.com/zh-HK/topics/100000000372028.md)
author: "[熊猫投zi手札](https://longbridge.com/en/profiles/25225811.md)"
generator: "portal-rs"
---

# Meta 收购 Manus 的根本原因是什么？为什么网上的质疑声这么大？

Most people got it wrong at first glance.

Everyone is saying: Manus is an Agent, a niche, a shell.

But honestly, Meta isn't after any of that.

Just think for a second from Zuckerberg's perspective:

What he's most anxious about isn't "having a flashy Agent product,"

but—the AI war, where Meta is lagging behind in data and real-world experience.

Manus is really only valuable for two things.

First, the engineering environment.

Not some demo in a PPT, but the real-world practice where users use it daily and encounter pitfalls. 147 trillion tokens, creating over 80 million virtual computers, serving millions of global users.

Models are smart in the lab but act dumb in the real world—that's exactly the missing piece for Meta right now. And Manus happens to fit all these criteria.

The second is even more critical: real agent traces.

Namely:

What users most often ask AI to do

Which tasks are most valuable

What tools are used

Which step is most prone to failure

These things can't be achieved overnight, even with $2 billion in funding.

They're built up over time through real user behavior.

So you say Manus is just a shell?

Doesn't matter.

Meta isn't after the shell—it's after the usage patterns and failure experiences behind it.

That's why:

Meta isn't buying a 2C business,

it's buying "the answer to training the next-gen model."

Who is Zuckerberg up against?

OpenAI, Google.

This isn't a product war anymore—it's an arms race.

Whoever figures out "what users really care about and where models suck the most" first will iterate their flagship model faster.

As for the money issue, the market is already hinting at it.

Meta's stock dipped slightly after hours, not because of Manus,

but because people have noticed one thing:

Meta is running a bit low on cash lately.

After Q3:

Directly issued $30 billion in bonds

Another $27.3 billion in "off-balance-sheet financing"

Capex keeps soaring

Cash + short-term investments on hand: only $44 billion

Down nearly 40% YoY

What does this mean?

It means Meta has entered a phase where:

👉 It'd rather be cash-strapped than wait any longer on AI accumulation.

One last emotional note.

The Manus deal is actually a morale booster for young entrepreneurs.

Not because "even a shell is valuable,"

but because:

If you truly stand on the side of real users and accumulate data and experience others don't have, someone will pay for it.

Zuckerberg is anxious about AI,

but at least he's still willing to believe in young people.

That kind of boss—the more, the better.

### Related Stocks

- [META.US](https://longbridge.com/en/quote/META.US.md)
- [OpenAI.NA](https://longbridge.com/en/quote/OpenAI.NA.md)
- [DXYZ.US](https://longbridge.com/en/quote/DXYZ.US.md)
- [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md)
- [GGLS.US](https://longbridge.com/en/quote/GGLS.US.md)
- [GGLL.US](https://longbridge.com/en/quote/GGLL.US.md)
- [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md)

---
> **Disclaimer: This article is for reference only and does not constitute any investment advice.**