China's Manus AI is being hyped as the next big leap in agentic AI—maybe even a "DeepSeek moment"—but early testers say the platform might be more sizzle than substance.
What Happened: Developed by Chinese startup Monica, the daughter company of The Butterfly Effect, Manus is billed as an "autonomous agent" capable of handling complex tasks like booking flights, ordering food, and even programming video games.
Its invite-only beta sparked massive buzz, with over 130,000 users joining its Discord server and invite codes reportedly selling for thousands of dollars on Chinese reselling app Xianyu, reported TechCrunch.
Some influencers and tech executives praised Manus as revolutionary. Hugging Face's head of product called it "the most impressive AI tool I've ever tried," while AI researcher Dean Ball dubbed it "the most sophisticated computer using AI."
But some user experience tells a different story.
Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox.
Alexander Doria, co-founder of AI startup Pleias, wrote on X, formerly Twitter, that he encountered "endless loops" and errors.
Zengyi Qin, a generative AI researcher, also took to X and shared his views saying, "Manus is NOT a technology breakthrough," because it mainly follows pre-defined workflows to perform tasks and operates within fairly limited environments.
Manus' website makes lofty claims—like buying real estate or building apps autonomously—but some of the platform's core functions reportedly rely on fine-tuned versions of existing models like Anthropic's Claude and Alibaba Group Holding Ltd.'s (NYSE BABA) Qwen.
Yichao "Peak" Ji, a research lead for Manus, called the tool a "completely autonomous agent" and claimed it outperforms OpenAI's Deep Research on the GAIA benchmark.
But a viral video touting Manus' capabilities was later confirmed by Ji to not be a real demo.
Why It's Important: The rise of Manus comes at a time when the AI landscape in China is rapidly evolving. DeepSeek, another Chinese AI firm, has been under scrutiny by the U.S. government, which is considering restrictions on its operations.
DeepSeek’s AI models have been made open-source, allowing them to be freely downloaded and copied.
In January 2025, DeepSeek’s assertion that its AI model was developed at a fraction of the cost compared to its U.S. competitors resulted in Nvidia Corp NVDA losing $600 billion in market value in just one day.
Moreover, a study by Copyleaks revealed that 74% of DeepSeek’s output mimics OpenAI’s ChatGPT, raising questions about its development process.
Meanwhile, earlier this month, Tencent’s Yuanbao AI has surpassed DeepSeek as the most downloaded iPhone app in China.
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Read Next:
Disclaimer: This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.
Photo courtesy: Shutterstock
Edge Rankings
Price Trend
© 2025 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.