Manus AI takes centre stage in China’s growing AI scene
”
A new AI system out of China is quickly making waves in the tech world. Named Manus, the product promises to take autonomous task execution to the next level, and some are already calling it a second “DeepSeek moment” for the AI industry.
Developed by the Chinese startup Butterfly Effect, Manus is more than just an AI model—it’s an interface for controlling multiple AI models that can autonomously complete complex tasks. This includes everything from generating reports to managing social media accounts, and more.
What is Manus?
Manus is a multi-agent AI system that can autonomously carry out tasks on behalf of its users. Unlike conventional chatbots that only respond to prompts, Manus is designed to take on tasks independently, such as generating reports, automating workflows, and even writing and deploying code. The system operates asynchronously, meaning users can assign tasks and walk away, while Manus works in the background.
The team behind Manus includes Xiao Hong, a serial entrepreneur, and several product managers. Hong previously launched Monica.ai, an AI assistant, and now leads Manus, which was officially introduced on March 5, 2025. It has already garnered significant attention in China’s AI ecosystem.
How does Manus work?
Manus is powered by a multi-agent system that combines several AI models to execute tasks. The system is built on Anthropic’s Claude 3.5 Sonnet, a nine-month-old AI model, and fine-tuned versions of Alibaba’s Qwen models. The team is also testing an upgrade to Claude 3.7 to improve Manus’ performance further.
The AI works asynchronously, allowing users to assign tasks and let Manus handle them autonomously. This sets it apart from traditional AI systems, which often require constant user input.
Manus AI’s rise in China and beyond
The buzz around Manus intensified after it was tested on freelance platforms such as Upwork and Fiverr, where it successfully executed complex real-world tasks.
AI influencers and early adopters have been particularly impressed with Manus’ ability to manage multiple tasks, such as creating websites, developing AI courses, and researching top rental spots in San Francisco.
Rowan Cheung, founder of the Rundown AI newsletter, called Manus a “game changer” and noted that it could combine the best features of other AI systems, such as Deep Research, OpenAI’s Operator, and Claude Computer. Early tests show Manus can handle tasks with exceptional efficiency, often outperforming competitors in benchmark tests.
How does Manus compare to US-Based competitors?
One of the key reasons Manus has gained attention is its benchmark performance. The system has outperformed OpenAI’s Deep Research agent and other state-of-the-art models, according to tests published on the Manus website. This has led some to believe that Manus could be the most capable autonomous AI agent currently available.
However, despite the positive performance, some have raised concerns about Manus’ reliance on existing AI models rather than creating its own foundation. Critics have questioned how much of Manus’s technology is truly original, especially when compared to DeepSeek-R1, which developed its own foundational model.
Manus sparks interest among influencers and early adopters
Manus’s launch has been met with enthusiasm in AI circles. Influencers like Bilawal Sidhu and Chubby have been particularly vocal about the tool’s capabilities. Sidhu, a former Googler and AI YouTuber, praised Manus for its ability to break down complex workflows and execute them autonomously.
The AI has also been tested for its ability to execute more routine tasks, such as generating reports, pulling data from social media sources, and researching locations based on specific parameters. Critics note that while Manus excels in some areas, it may still be lacking in other domains, particularly in STEM applications.
Questions about Manus’ future
Despite the excitement, there are concerns about Manus’s scalability and availability. Currently, Manus is only available to select users through an invite-only beta, and the team behind the tool has admitted that server shortages have led to frustration among potential users.
Invitation codes for Manus have even been sold on China’s second-hand marketplace, Xianyu, adding to the hype but also leading to criticisms of the product’s exclusivity.
Additionally, while Manus is praised for its ability to execute tasks, some believe that it may be overhyped. Critics point out that it remains unclear whether Manus can replicate the success of other AI products like DeepSeek, especially in the long term.
Manus AI and content control
Manus has shown to provide uncensored responses to sensitive topics, including historical events like the Tiananmen Square protests, which other AI systems like DeepSeek have avoided.
This has raised questions about Manus’s content control mechanisms, especially in a country with strict internet censorship laws. However, the Manus team maintains that it does not intentionally censor factual information and that its focus is on task execution rather than content regulation.
Future of Manus Ai
While Manus has made an impressive debut in China’s AI landscape, its true potential remains unclear. The system’s ability to scale and handle more complex tasks could make it a major player in enterprise automation.
However, as Manus faces server limitations and access restrictions, it’s uncertain whether it can achieve widespread adoption.
Some analysts suggest that Manus could be a breakthrough in human-machine collaboration, but others remain sceptical, questioning whether the system can maintain its current momentum.
”
