What happened
Chinese-founded AI agent company Manus is close to closing a roughly $500 million funding round at a target valuation of about $4 billion — double its previous mark, according to Bloomberg. People familiar with the matter say terms could still change and the new investors' identities are not yet clear.
It is Manus's first round since regulators forced it to split from Meta. Completed as planned, it would make Manus the most valuable agentic-AI startup in China.
Background: a deal that was unwound
The timeline is short. In December 2025, Meta announced its acquisition of Manus for about $2 billion after the startup passed $100 million in annualized revenue. In April, China's foreign-investment security review office blocked the deal, citing concerns that China risked losing valuable technology to a geopolitical rival. The two companies completed their operational split and halted all data-sharing in May.
In July, the founding team and existing backers Tencent, HSG, and ZhenFund bought Meta's stake back at the same $2 billion valuation; Tencent became the largest external investor after Benchmark exited. Earlier this month, Manus said it had resumed independent operations with the founding team still leading. In August, it notified some users that their data would be deleted to comply with regulatory requirements.
Why this round matters
It is both a blueprint and a cautionary tale. Manus shows how a China-founded startup navigates Silicon Valley capital, an American acquirer, and Chinese regulation at once: a $75 million Benchmark-led round at a $500 million valuation in April 2025; acquired and then forcibly unwound; now back at $4 billion.
The round also signals sustained appetite for the agent category: rival Evoken, behind AI design agent Lovart, is raising at a $3 billion valuation. But Manus does not train its own base models from scratch, leaving it vulnerable to general-purpose models — Moonshot's Kimi, Anthropic's Claude — that increasingly handle desktop tasks as well.
Whether the $4 billion holds depends on the round actually closing and on new capital turning demos into a business that earns its keep. Agent startups do not lack narratives; they lack models with math that works.
[1][2]