Cover Image | Produced by Pencil News

In September, Manus once again made headlines with financing rumors. Sources familiar with the matter said the company plans to raise approximately $500 million, targeting a valuation of around $4 billion (approximately RMB 27.8 billion), and is considering restructuring in preparation for a future Hong Kong IPO.

This marks Manus's first financing round since its spin-off from Meta and restoration of independence.

Late last year, when Meta acquired Manus, the deal was valued at around $2 billion. In August this year, Manus's existing shareholders, including Tencent, ZhenFund, and HSG, bought it back from Meta for approximately $2 billion.

The transaction was revoke (revoked, corrected to 'revoked') , and Manus's valuation doubled, but the controversies surrounding it are far from over.

- 01 - A Company Skilled in Marketing

Manus first gained popularity amid marketing controversies.

When the product was launched in March 2025, it adopted an invitation-only system. Within days, the official Discord community had surpassed 138,000 members, and some invitation codes began to be resold at high prices on platforms like Xianyu, with some codes priced at tens of thousands of yuan.

An investigation by National Business Daily found that after Manus's release, a large number of Chinese self-media outlets quickly took notice, using particularly exaggerated language—stunning the world, achieving a national-level accomplishment. Many questioned whether Manus was a technological marvel or a carefully orchestrated scarcity hype.

However, Manus explained that the invitation-only system was due to insufficient server capacity.

The shortage of invitation codes gave Manus tremendous exposure: the harder the product was to obtain, the more people discussed it; the more expensive the invitation codes, the stronger the impression that 'this product is worth snapping up'; the prices in the secondary market continued to make headlines in media reports.

Two weeks later, at an event in San Francisco, Manus disclosed that its waitlist had exceeded 2.6 million people, and its X account had over 150,000 followers.

However, early testing by TechCrunch found that Manus could complete complex tasks such as web research, report writing, and script execution, but some tasks reported errors after running for about half an hour. A video allegedly showing Manus operating multiple mobile apps also circulated on social media, but co-founder Ji Yichao later confirmed that it was not a demonstration of Manus. TechCrunch's initial assessment was that Manus's hype had outpaced its technological innovation.

Even so, the traffic quickly translated into revenue. In late March, Manus introduced two paid plans priced at $39 and $199. By August 2025, Manus's annualized revenue run rate had reached $90 million; in December, the company announced that its ARR had surpassed $100 million. From product launch to $100 million ARR, it took only about eight months.

In April this year, an investigation by tech media outlet The Verge took Manus's growth strategy to another level.

By that time, Manus had already been acquired by Meta. The Verge discovered that Manus was running paid UGC creator programs on platforms like TikTok, Instagram, and YouTube, with a large number of young creators posting highly similar content, focusing not on Agent technology but on 'making money.'

One common script involved using Manus to find local small businesses without websites or with poorly designed ones, having AI generate a website in minutes, and then selling the website to the business.

Some videos portrayed it as a side hustle anyone could do, claiming it could be completed in under 10 minutes with a monthly income potential of up to $5,000.

The Verge found that these accounts were highly similar in visual style, messaging, and even video themes. One participant described themselves on LinkedIn as a 'viral growth expert' hired by Manus, leading a team of 10-20 content creators and training them to produce viral content according to unified brand guidelines.

Manus did not deny the program.

A company spokesperson confirmed to The Verge that Manus was indeed collaborating with creators on TikTok, Instagram, and YouTube through third-party agencies on paid UGC projects.

The controversy lay in the fact that some accounts did not clearly disclose their commercial relationship with Manus.

Advertising and legal experts interviewed by The Verge believed that such undisclosed commercial collaborations might violate advertising policies on some platforms and could even breach advertising regulations in certain jurisdictions. After inquiries from the report, some TikTok videos were deleted, and a batch of related accounts became inaccessible.

Another issue was the earnings promises.

Manus told The Verge that the company 'does not endorse exaggerated or misleading earnings claims' and was reviewing the content in question. However, it did not provide specific evidence for figures like '$5,000 per month.'

This small advertising incident continued Manus's characteristic approach since its inception: highly skilled in marketing and growth.

- 02 - Controversy Over 'Wrapper'

When Manus first gained popularity, a common question was: Is this really a tech-savvy AI company?

In March 2025, just days after Manus's release, accusations of it being a 'wrapper' emerged. Manus did not train its own foundational large model; its core reasoning capabilities came from Anthropic's Claude, while also using Alibaba's Qwen model.

In a report, one user summarized Manus as 'not offering much new technology but integrating well.' An industry expert in large models also judged that while Manus's product and engineering capabilities were strong, its technological concept was not new, similar to the Auto-GPT approach that had emerged two years earlier.

The 'wrapper' accusation captured part of the truth but oversimplified Manus. Manus built a functional system around existing large models and made other models work stably and continuously. Silicon Valley 101 later noted in its review of Manus that the real challenge in Agent technology is not just connecting a few tools but ensuring reliability. In a task involving dozens of steps, small errors at each step can accumulate, leading to the failure of the entire task.

Li Mingshun, Chairman of Hanghang AI and Founder of Shunfu Capital, offered a judgment: 'The core issue is not whether it's a wrapper.'

In his view, as large models become increasingly mature, foundational capabilities will gradually converge. The real question becomes: Can you create a product that users are willing to use and pay for, and then gradually accumulate users, data, branding, and workflows?

A typical example this year is Harvey.

Harvey, a legal AI company, also did not train a general-purpose foundational model, but this did not prevent capital from investing.

In September, Harvey secured $550 million in financing, reaching a valuation of $15.5 billion. The company disclosed that 80% of the U.S. Am Law 100 law firms were using Harvey, and its clients also included legal departments of multiple Fortune 500 companies.

Lawyers don't really care which large model you use behind the scenes. What they care about is the result: whether a 100-page contract can quickly identify risks; whether thousands of documents can be automatically organized; whether tasks that used to take young lawyers several all-nighters can be delivered by AI in a few hours.

What Li Mingshun values about Manus is precisely this point.

In his view, Manus relatively early seized the opportunity in Agent workflows and attempted to transform AI from a 'tool for chatting' into a product that could truly perform tasks and deliver results for users.

'AI used to tell you how to do things; in the future, AI will do the work itself.'

- 03 - Manus Becomes 'Cheaper'

Manus's valuation doubled, but when revenue is factored in, it has actually become 'cheaper.'

At first glance, the price seems to have risen rapidly.

In April 2025, when Benchmark invested $75 million, Manus was valued at approximately $500 million; eight months later, Meta acquired it for around $2 billion; now, with the transaction halted by Chinese regulators and Manus newly independent, the market is preparing to raise its valuation to $4 billion.

In one and a half years, the valuation has increased sevenfold.

However, when revenue is considered, there is another interpretation.

When Meta acquired Manus last December, the company had just announced that its ARR had surpassed $100 million, with total revenue annualized run rate exceeding $125 million. Based on the $2 billion acquisition price, this corresponded to approximately 20 times ARR.

A few months later, revenue rapidly increased.

Caixin reported that as of the end of June this year, Manus's ARR had reached approximately $400 million, four times that of last December; The Information, citing sources familiar with the matter, put the figure even higher, with the most recent annualized revenue run rate reaching $400-500 million.

Based on these figures, the income multiple for the $4 billion valuation is only about 8-10 times.

In other words:

Manus's valuation doubled from $2 billion to $4 billion, but its revenue increased from approximately $100 million to $400-500 million, growing 3-4 times.

If these revenue figures hold, Manus is actually 'cheaper' today than when Meta acquired it.

This change became apparent two months ago.

After Meta was required to revoke the transaction, investors including Tencent, ZhenFund, and HSG bought Manus back for around $2 billion. At the time, some media outlets calculated: based on an annualized revenue of nearly $500 million, the income multiple for this buyback was only about 4-5 times, whereas Meta had paid approximately 20 times just half a year earlier. The media outlet thus argued that while the original investors were taking over under regulatory requirements, what they bought back was not the same company as six months ago.

This may be good news for investors, who are also betting on three other things: that the current $400-500 million in revenue can continue to grow; that users were not just temporarily attracted by a marketing campaign; and that after paying for Claude, servers, and customer acquisition costs, Manus can still maintain a sufficiently high gross margin.

This article does not constitute any investment advice.