Manus Raises More Than $500 Million After Meta Deal Falls Apart

Butterfly Effect, the parent company of Chinese AI lab Manus, announced on October 8, 2026 that it raised more than $500 million, according to TechCrunch. The round is the company's first since its $2 billion acquisition deal with Meta was called off.
Boyu Capital and IDG Capital led the financing. Tencent, HSG, formerly Sequoia China, and ZhenFund took part. The company did not disclose a valuation.
That leaves one earlier figure as public but unconfirmed. In September 2026, Manus was reported to be in talks to raise $500 million at a $4 billion valuation. The announced total came in above $500 million, with no valuation given.
The financing follows a run of events for the lab. Manus moved its staff to Singapore in mid-2025. It announced the $2 billion acquisition deal with Meta in December 2025. At the time of that deal, the company had more than $100 million in annual recurring revenue, or ARR, the yearly income from ongoing subscriptions and contracts.
In April 2026, Chinese authorities ordered Manus to unwind the Meta deal. Manus resumed independent operations in August 2026 after the deal ended. As part of the split, Manus said it was required to delete some user data.
On the ownership side, the setting is important. The lead and participating firms are China-based investors. For a lab that had just left a U.S. acquisition agreement on regulatory orders, a domestic group gives a cleaner legal position than a mixed cross-border group. It also puts control among investors used to working under Beijing's rules on outbound investment and data.
The broader context here is operational, not just financial. An unwind that requires deletion of some user data is costly in ways that do not appear in a funding headline. Deletion affects training pipelines, the systems that teach models, evaluation sets, the tests used to check them, logging infrastructure, retention policies and customer trust. A lab with more than $100 million in ARR at the time of the Meta agreement has live customer workloads to keep running through that change. Fresh capital restores an independent balance sheet. It does not automatically bring back datasets or product momentum.
Looking at what this means for Manus as an operator, the missing valuation matters. Holding back a valuation after the reported $4 billion September target avoids setting expectations while the company rebuilds post-separation numbers. Business buyers and API customers watch ARR, churn, or customer loss, latency, or response speed, and reliability more closely than headline valuations. For technologists judging the stack, the questions are narrower. What was deleted, what was kept, and how the Singapore entity now handles residency, access controls and audit, or where data lives, who can touch it, and how it is checked.
In my view, the August resumption of independent operations and the October raise should be read together. The first restored control. The second pays for the cost of using it. Labs that come out of a failed acquisition often end up with stricter data hygiene and clearer ownership of infrastructure than they had before the deal. That is unglamorous work. It is also the work that decides whether revenue can grow a second time without the acquirer's sales reach and computing resources.
Longer term, the reason for optimism is simple. Demand for applied AI systems did not pause while ownership was in dispute. A team that kept nine-figure recurring revenue into a $2 billion deal, then raised more than $500 million to continue alone, has room to rebuild systems on its own terms. The test now is execution under tighter limits.


