Manus Raises More Than $500 Million After China Blocks Meta Deal

Preface

Manus, an AI agent startup, has raised more than $500 million in a new funding round after a proposed acquisition by Meta was blocked by Chinese authorities. The financing marks a significant turn in the company’s story: it was acquired, ordered to separate from its buyer, and then returned to independent operations. Its parent, Butterfly Effect, confirmed the investment in a WeChat post. Boyu Capital and IDG Capital led the round, while existing backers Tencent, HSG and ZhenFund also participated. The company did not disclose a valuation or explain in detail how it will use the new capital. The funding follows rapid commercial growth, an international relocation, and growing competition in AI agents—software designed to complete tasks rather than simply respond to prompts. The events also illustrate how cross-border investment and regulatory decisions can reshape the future of a technology company.

Lazy bag

Manus’s fundraising comes after China ordered Meta’s roughly $2 billion acquisition to be unwound, and the startup announced it would operate independently again. Its parent, Butterfly Effect, raised more than $500 million, with Boyu Capital and IDG Capital leading the round. Manus says it reached $100 million in annual recurring revenue about eight months after launching in March 2025, though the latest funding round’s valuation was not disclosed. The company plans to continue hiring in China and abroad as it develops AI agents and its Cue app, in a market increasingly shaped by competition, regulation and international expansion.

Main Body

Manus is an AI agent startup whose parent company, Butterfly Effect, has raised more than $500 million in fresh funding. Boyu Capital and IDG Capital led the round, with Tencent, HSG—the firm formerly known as Sequoia China—and ZhenFund participating as existing backers. Butterfly Effect announced the financing in a WeChat post. It did not disclose a valuation for the round, nor did it provide a detailed breakdown of how the capital will be spent. The company said it intends to continue hiring in China and abroad.

The new funding arrives after a proposed sale to Meta became one of the defining events in Manus’s recent history. Meta announced in December 2025 that it would buy Manus for roughly $2 billion. China’s National Development and Reform Commission, the country’s top economic planning agency, ordered the transaction withdrawn on April 27. Meta cut ties in June, and Manus announced in August that it would resume operating independently. The sequence left the startup with a new financial backer group and a renewed obligation to establish itself outside Meta.

Butterfly Effect did not reveal the company’s valuation in its announcement. Bloomberg had previously reported that the company was considering a $500 million raise at a $4 billion valuation, roughly double the amount Meta had agreed to pay. That reported target should be distinguished from the confirmed terms: the current round exceeded $500 million, but no valuation was disclosed. The available information therefore establishes the size of the financing while leaving the company’s post-funding value unknown.

Manus develops AI agents, a category of software intended to pursue a goal through multiple steps. Unlike a conventional chatbot, which primarily generates responses to user questions, an agent can be asked to carry out a task, such as booking a trip or analyzing stocks. The distinction is not that an agent never uses conversation; rather, its purpose is to move from instructions toward an outcome, using software tools and intermediate actions along the way. This approach has drawn interest from technology companies seeking to make AI useful for practical, multi-step work.

Manus launched in March 2025 and presented itself as a Chinese competitor to an agent from OpenAI that cost $200 a month. Access to Manus was initially invitation-only. Reports described intense interest in access codes, with some listings on a Chinese resale marketplace reaching as much as 10 million yuan, or over $1.3 million. These resale listings indicated significant public attention, though they were not a measure of the company’s revenue or the number of paying users.

The company’s early technical approach combined its own work with models from other providers. Manus acknowledged using Anthropic’s Claude and fine-tuned versions of Alibaba’s Qwen models alongside its own technology. That disclosure placed the startup within a broader pattern in which AI products can be assembled from a mix of proprietary software and external models. Such arrangements can help a company build a product, while also making its technical and commercial position dependent in part on model providers and the terms governing access to their systems.

Manus gained attention before AI agents had become a central focus for the largest technology companies. It was experimenting with agent-based tasks before major AI companies began investing heavily in similar products. The competitive landscape subsequently changed. Attention shifted toward OpenClaw, an open-source agent that runs on a user’s own machine and accepts instructions through apps such as WhatsApp and Telegram. Within weeks of going viral, OpenClaw had collected well over 100,000 GitHub stars, a measure of developer interest in the project. OpenAI later hired OpenClaw’s creator, Peter Steinberger, to lead its push into personal agents.

Manus also experienced substantial changes to its operations and geographic footprint. Around mid-2025, it moved its team to Singapore, shut most of its China operations and laid off dozens of employees in July. Those changes took place as the business was expanding internationally and as questions about the acquisition and its regulatory treatment became more prominent. The company’s later plan to keep hiring in China and abroad suggests that its new phase will involve maintaining a presence across multiple markets, even after it had reduced its earlier China operations.

Despite the upheaval, Manus reported reaching $100 million in annual recurring revenue by December, about eight months after its March 2025 launch. Annual recurring revenue, commonly abbreviated as ARR, is an annualized measure of recurring subscription income. It is not necessarily the same as cash collected or total recognized revenue over a year. The figure offered evidence of commercial traction, but the company did not provide further detail about its customer base, the composition of subscriptions or how the metric would change after the acquisition was unwound.

In December, Meta announced its plan to acquire Manus for roughly $2 billion. The deal became subject to scrutiny from Chinese authorities the following month. China’s commerce ministry said in January that it would assess the transaction. By March, Manus co-founders Xiao Hong and Ji Yichao had been summoned to Beijing and barred from leaving the country, according to Reuters. On April 27, the National Development and Reform Commission ordered the deal withdrawn and said it would “prohibit foreign investment in Manus in accordance with laws and regulations.” Meta cut ties in June. In August, Manus said it would operate independently again and deleted some user data created on or after December 29, 2025, as part of separating its systems from Meta’s.

At the time Manus was returning to independence, competition in agent software was intensifying. Meta’s coding agent, Muse Code, launched in August into a market where it appeared to be behind products including Codex and Claude Code. Manus, meanwhile, now sells Cue, an app that gives agents their own phone numbers and digital wallets. Payments through Cue are limited to a budget set by the user, a design that places a spending boundary around actions an agent may take. The product reflects one approach to making agents more capable while giving users a way to constrain financial activity.

The Manus episode also fits into a wider tightening of oversight in China. In May, China required some senior AI workers at private firms, including Alibaba and DeepSeek, to obtain approval before traveling abroad. The policy arrived as Chinese AI systems approached levels of quality and relevance that could compete directly with American companies. Taken together, the restrictions on travel, scrutiny of a foreign acquisition and the unwinding of Meta’s agreement show how government policy can influence the ownership, movement and operations of companies working in strategic technology sectors.

For Manus, the immediate question is how it will use its new capital to build an independent business after the abandoned sale. Its rapid reported ARR growth and investor support provide a foundation, but the company has not disclosed its valuation or detailed spending plans. It must also navigate competition from both established AI companies and open-source projects, while deciding how to balance its operations in China and abroad. The fundraising is therefore an important milestone, but it does not resolve the strategic and regulatory uncertainties that have shaped the company’s path.

Key Insights Table