China’s Manus raises over $500M in first funding round since split with Meta

The $500 Million Gamble on Autonomy

We are past the point of being impressed by simple chat interfaces. The industry is shifting from bots that talk to agents that do. Manus, a startup that recently emerged from the shadows of its previous association with Meta, just landed over $500 million in a funding round led by Boyu Capital and IDG Capital. This isn't just another AI raise; it is a signal that the infrastructure for autonomous agency is becoming a sovereign priority.

When a company pulls this kind of capital right after a split from a Western tech giant, you have to look at the motivations. Tencent and Sequoia China—now known as HSG—are doubling down here. They aren't just buying another LLM wrapper. They are betting on a future where AI handles complex, multi-step tasks without a human holding its hand every five seconds.

Why the Split Matters

Manus leaving the Meta ecosystem is the headline, but the subtext is about independence. In the current geopolitical climate, being tethered to a Silicon Valley stack is a risk for any founder building in Asia. By raising half a billion dollars, Manus is effectively buying its own runway to build a stack that doesn't rely on the permission of Western gatekeepers. For builders, this is a lesson in strategic decoupling. If your core product depends on an API that can be throttled by a foreign government, you don't really own your business.

The participation of ZhenFund and existing shareholders shows a rare alignment of old-guard venture capital and new-world aggressive scaling. They see the same thing I do: the low-hanging fruit of AI—generating text and images—is already commoditized. The real value is in agency.

The Founder Perspective: Scaling Without Bloat

I have seen plenty of founders raise nine figures and immediately light it on fire by hiring 500 middle managers. The challenge for Manus is to remain a builder-first organization while managing a war chest that large. When you have $500 million, the temptation is to solve every problem with money rather than engineering elegance.

For those of us building in the trenches, the Manus raise is a double-edged sword. On one hand, it validates the market for autonomous agents. On the other, it creates an environment where talent costs are going to skyrocket. If you are a founder trying to hire agentic workflow engineers right now, you are officially competing with a company that has more cash than some small countries.

The Shift to Agentic Workflows

What exactly is Manus building? While the specifics are often wrapped in PR speak, the focus is clearly on the leap from "Copilots" to "Autopilots." A Copilot waits for you to type. An Autopilot sees a goal, breaks it down into sub-tasks, and executes them across different software environments. This requires a massive amount of compute and even more sophisticated error-correction loops.

The real technical hurdle isn't making the AI smart; it is making the AI reliable when it encounters a UI change or a broken API.

Manus is likely spending a significant portion of this $500 million on the compute necessary to train models that understand environment navigation. Most LLMs are trained on text. Agents need to be trained on actions. That is a much more expensive data set to acquire and process.

What This Means for the Crypto/AI Intersection

If you are building in the decentralized AI space, you should be watching Manus closely. Autonomous agents will eventually need a way to transact. They won't have credit cards or bank accounts in the traditional sense. They will need programmable money. This is where the crypto side of the house meets the agentic side. A $500 million agent player is a massive potential customer for on-chain identity and payment rails.

However, I remain skeptical of the timeline. We have seen "autonomous" hype before. The difference now is the sheer volume of capital being thrown at the problem. Money doesn't always solve logic errors, but it does allow you to fail faster and iterate more times than the competition.

The Takeaway for Builders

Don't get distracted by the $500 million number. Most of that will go to NVIDIA or its equivalent. Instead, look at the geography and the backers. The East is building its own AI sovereign state. If you are a founder, you need to decide if you are building for one ecosystem or trying to bridge both.

The era of the simple chatbot is over. We are moving into the era of the autonomous worker. Whether Manus succeeds or becomes a cautionary tale of overfunding, they have officially fired the starting pistol for the agent arms race. Keep your teams lean, focus on specific workflows, and don't try to out-spend these guys. Out-innovate them by finding the niches their massive models are too heavy to touch.

Read the original at TechCrunch Startups →