Manus, the AI agent startup whose acquisition by Meta collapsed under pressure from Beijing earlier this year, is in discussions to raise $500 million at a $4 billion valuation — roughly double the price at which its early backers bought the company back from its would-be acquirer.

The round, first reported by the Wall Street Journal citing anonymous sources, would be Manus’s first major financing as an independent company. It also doubles as a market test of a simple question: in an industry that keeps consolidating agent startups into big tech, can a mid-sized agent company fund its way to independence?

The deal that died#

Meta agreed to acquire Manus for about $2 billion in December 2025, shortly after the startup moved most of its staff to Singapore. At the time, Manus was reportedly pulling in more than $100 million in annual recurring revenue — a substantial figure for an AI company its age.

The deal never closed. Growing anxiety in China about AI talent and research flowing to the West led Beijing to block the transaction, with regulators citing potential violations of export controls and foreign investment rules. According to one account of the episode, China’s National Development and Reform Commission ordered the deal unwound in April, on national-security grounds.

What followed was a slow, expensive disentanglement. Early investors helped Manus buy back its shares at a valuation of around $2 billion. The companies completed their operational separation and halted data-sharing; in August, Manus told users to export and back up their own data because it had to delete information generated after Meta’s acquisition to satisfy regulatory requirements in specific jurisdictions. This month, the company said it has resumed independent operations, with its founding team still in charge.

Who’s writing the checks#

The Journal’s sources name IDG Capital, Boyu Capital, and battery maker Contemporary Amperex Technology as prospective new investors, alongside existing backers Tencent, HSG, and ZhenFund. Manus is also reportedly considering a restructuring to prepare for an initial public offering in Hong Kong. Terms remain fluid, the round has not closed, and Manus did not return a request for comment.

One notable detail: Tencent may emerge from the process as Manus’s largest external shareholder, after acquiring the stake previously held by Benchmark — an early investor that reportedly exited with a multi-fold return when the Meta deal was unwound. The investor mix is telling: classic venture firms, a domestic tech giant, and a battery manufacturer — the last suggesting agents are being read as industrial infrastructure, not just apps.

What Manus actually sells#

Manus went viral last year after a demo of its AI agent, and its pitch has stayed consistent: general-purpose agent software that handles chat, builds apps and websites, produces designs and presentations, and generates video. It competes in the same lane as OpenAI’s agent offerings and startups like Lovable and Replit — the crowded market for tools that turn natural language into working software.

The revenue figure matters here. More than $100 million in annual recurring revenue at the time of the Meta deal is the kind of traction that separates a viral demo from a real business, and it is presumably the foundation the $4 billion valuation claim rests on. Whether that number has grown, held, or slipped during a year spent unwinding a merger is the question every prospective investor will be asking.

Why this raise matters#

Three threads make this story bigger than one startup’s cap table.

First, geopolitics. Manus is becoming the reference case for what happens when an AI deal crosses a geopolitical fault line. Beijing treated AI talent, data, and model technology as strategic assets and killed a $2 billion transaction to keep them. That is now a risk premium priced into every cross-border AI deal involving Chinese-founded teams — and a reason more of them will be financed onshore instead.

Second, the independence bet. The dominant story of the agent market this year has been consolidation: startups absorbed into frontier labs and hyperscalers, teams folded into product lines. A $4 billion standalone valuation, if it holds, is the strongest counterargument available — evidence that investors still believe an agent company can build its own distribution and enterprise traction without a parent.

Third, the arc: backers bought the company back at roughly $2 billion, and targeting $4 billion now is a claim the business is worth twice what it was at its lowest point. Manus paid a high price to stand alone; this round is where it starts trying to earn it back.

What to watch#

Watch whether the round closes at $4 billion — the reporting stresses talks are ongoing and terms fluid — and whether the Hong Kong IPO restructuring materializes. And watch revenue: the $100-million-plus ARR figure dates to December, and growth through a year of upheaval would make the $4 billion number look like a floor, not an aspiration.