Manus, a New AI Agent From China is Going Viral—And Raising Big Questions

A new team out of China has launched Manus, a general AI agent that has rapidly become the talk of the global AI community. The product is already drawing widespread comparisons to last month's viral DeepSeek moment, with some industry leaders calling it 'the AI agent we were promised.' But, with invitation codes in short supply, and demand through the roof, getting access anytime soon won't be easy.

Key Points:

• Manus functions as a truly autonomous agent, independently analyzing, planning, and executing complex tasks from start to finish
• Early testers report it completes weeks of professional work in hours, with authentic agentic capabilities that go beyond existing AI assistants
• The system uses a multi-agent architecture, breaking down complex tasks into components handled by specialized sub-agents

So, why the hype? Well, Manus operates independently—analyzing, planning, and executing tasks on its own. And it seems to do this really well. Early testers are calling it the first true autonomous AI agent, capable of doing weeks of work in hours. And with prominent voices from the AI community raving about its capabilities, many are now asking: Has China leapfrogged the US in AI, and how will the western labs respond?

"While other AI stops at generating ideas, Manus delivers results," said Peak Yichao Ji, co-founder and chief scientist of Manus, in the official demonstration video. "We see it as the next paradigm of human-machine collaboration and potentially a glimpse into AGI."

During its launch, the company demonstrated capabilities across diverse tasks: screening resumes, researching real estate properties based on multiple criteria, and performing correlation analysis between stocks—all with minimal human intervention.

What's particularly impressive is Manus's reported performance on the GAIA benchmark, a comprehensive AI test developed by Meta AI, Hugging Face, and the AutoGPT team that evaluates an AI's ability to reason, use tools, and automate real-world tasks.

The reaction from industry leaders has been swift and enthusiastic. Clement Delangue, Co-founder and CEO of Hugging Face, shared insights from conversations with Manus co-founder Yichao Ji, noting that "agentic capabilities might be more of an alignment problem rather than a foundational capability issue."

This suggests Manus may have achieved its impressive capabilities through innovative fine-tuning rather than core model innovation. The system's technical architecture operates as a multi-agent system that divides complex tasks into manageable components and assigns them to specialized sub-agents.

Venture capitalist Deedy Das was equally impressed, calling Manus "the AI agent we were promised" and detailing how it completed what would normally be "two weeks of professional-level work in about one hour" when asked to analyze Tesla stock.

Victor Mustar, Head of Product at Hugging Face, described Manus as "the most impressive AI tool I've ever tried," highlighting its mind-blowing agentic capabilities and user experience that "just works" where others have only promised.

Despite its impressive capabilities, Manus remains difficult to access due to limited server capacity, creating a scramble for invitation codes. "The current invite-only mechanism is due to genuinely limited server capacity at this stage," Zhang Tao, Manus AI's product partner, explained in a social media post. "The current version of Manus is still in its infancy, far from what we aim to deliver in our final product."

While some have questioned whether Manus represents genuine technological innovation since it builds on existing large language models, investor Andrew Wilkinson offered perhaps one of the most striking endorsements: "I just got access to Manus, the Chinese AI agent everyone is raving about. It's absolutely insane. I feel like I just time travelled six months into the future."

With endorsements like these from top AI researchers, venture capitalists, and industry leaders, Manus appears to be living up to its promise as a true autonomous agent rather than just another incremental improvement. As Rowan Cheung succinctly put it: "I think China's second DeepSeek moment is here."

Calling this a definitive shift in AI power dynamics may be premature, but one thing is clear: Manus and DeepSeek have commanded the attention of the global AI community like few products in recent memory. These back-to-back breakthroughs signal that U.S. dominance in AI is no longer a given—China is now a serious contender in the race for AI supremacy. Western labs will have to accelerate their own agentic AI development, or risk being left behind.