From Wuhan to Silicon Valley, Manus took nine months

Author: Frog Frog, Deep Tide TechFlow

Today’s biggest news in the AI circle: Meta acquires Manus for billions of dollars.

This is Meta’s third-largest acquisition in history, only behind WhatsApp and Scale AI, and more expensive than the Instagram acquisition back then.

Looking at Manus’s timeline, it’s clear that the product only launched in March of this year, and was acquired in December. From release to sale, a total of 9 months.

Founder Xiao Hong, from Ji'an, Jiangxi, graduated from Huazhong University of Science and Technology, with startup roots in Wuhan. The first product was a typesetting tool called YiBan, which was sold. The second was an enterprise CRM tool called WeiBan, also sold. The third was a browser AI plugin called Monica, which wasn’t sold but was criticized.

Criticized for what? For shelling.

At that time, the industry consensus was that only companies developing large models had a future; applications built on others’ models were just shelling, with no technical content.

When Manus was just gaining attention in March this year, co-founder Ji Yichao responded to a question on social media, saying: “We’ve used Claude, and also fine-tuned different versions of Qwen (Qianwen).”

They used others’ large models to build their applications.

So what?

Now valued at billions of dollars.

Last year, ByteDance executives flew to Hong Kong to meet Xiao Hong, offering $30 million to acquire. Xiao Hong didn’t sell.

Looking back now, the difference between $30 million and billions isn’t just a year — it’s:

They built a product.

Moreover, we think the most interesting part of this story isn’t the ending, but the process.

In July this year, Manus made a decision: move the company from China to Singapore. The team of 120 people kept only 40 core technical staff, and laid off the rest. The Beijing office was closed, and the Wuhan office was also shut down.

Many people criticized them as “greedy.”

Now, looking back, this step was necessary. It’s almost impossible for a Chinese company to get approval for acquisition by a US tech giant in today’s environment. Changing the registration location removes the obstacle.

The negotiation only took 10 days.

ZhenFund partner Liu Yuan said, at first, they even doubted whether this was a fake offer.

A multi-billion dollar deal closed in 10 days — how urgent was Meta?

Looking at the background: this year, Meta’s capital expenditure on AI exceeded $70 billion, but most of it was spent on infrastructure, with few products available for use. OpenAI has ChatGPT, Google has Gemini, what does Meta have?

Llama is open source, anyone can use it. Meta needs a competitive application-layer product, and Manus just happens to be ready.

Annual revenue of $125 million, built from scratch in 8 months, with global users, subscription model, capable of running smoothly.

This isn’t just about acquiring a team; it’s about acquiring a:

Proven business model.

Interestingly, Manus’s investors include Sequoia China, Tencent, and ZhenFund. When these investments were made, the valuation was in the tens of millions of dollars. Now, the exit yields dozens of times return.

So, you see, Chinese VCs invested in a Chinese company, which moved to Singapore, and was acquired by an American company — Chinese VCs profited from the US company’s money.

This chain is even more agent than Manus’s product.

After the acquisition, Xiao Hong became Vice President of Meta. An entrepreneur who started in Wuhan, made a typesetting tool, now reporting to Zuckerberg in Silicon Valley.

ZhenFund’s Liu Yuan said: “The era belonging to China’s young entrepreneurs has arrived.”

This statement might only be half correct.

The era has indeed arrived, but the way it arrives is by moving companies away.