Manus, the newly launched artificial intelligence (AI) agent from Butterfly Effect, has quickly caught the attention of the global technology sector by demonstrating the ability to efficiently complete tasks traditionally executed by white-collar professionals. Its debut on March 5, 2025, sparked intense interest, particularly across social media platforms where audiences showcased the agent’s diverse capabilities ranging from trip planning to financial analysis.
The Beijing-based startup, Butterfly Effect, has positioned Manus as the world’s first fully autonomous AI agent. Peak Ji, the chief scientist and co-founder of the company, revealed during the launch how Manus can swiftly perform various tasks. “For example, it can plan your trip to Japan or find new homes overseas—a feat completed within mere seconds,” he stated, highlighting the software’s efficacy.
Manus operates through advanced algorithms developed using models from Antropic’s Claude and Alibaba's Qwen, providing it with superior coding abilities. According to the company, the integration of these models allows Manus to function as more than just another chatbot or workflow tool. “Manus is a general AI agent intended to turn your thoughts directly to actions,” Ji remarked.
The recent announcement included the establishment of a strategic cooperation agreement between Alibaba Cloud and Butterfly Effect, aimed at advancing AI product development. This collaboration is expected to bolster Manus’s capabilities and applicability to various industries.
Promotional demonstrations from the launch displayed Manus's performance, such as its ability to analyze ten resumes within moments, search New York property listings under specified budgets, and calculate stock correlations for companies like Nvidia and Taiwan Semiconductor Manufacturing Co. It has reportedly surpassed OpenAI's Deep Research on the GAIA benchmark, achieving 86.5% accuracy compared to OpenAI’s 74.3%.
With such strong market performance, it’s no surprise the program has captivated users. Invitations for Manus's beta testing phase have sold for upward of $14,000 on secondary markets, indicating high demand and significant engagement. Despite this rising enthusiasm, Zhang Tao, co-founder of Butterfly Effect, acknowledged the unexpected amount of interest, stating, “We have overloaded our servers and sincerely apologize for the inconvenience.”
Manus embodies the potential shift from research-centered AI to practical application-focused solutions, diverging from approaches pursued by Western companies. Many compare it to previous notable AI innovations from China, like DeepSeek, which also experienced notable performance surprises and industry buzz. Observers argue Manus signifies China's ambition: the development of efficient AI tools prioritizing immediate usability over lofty theoretical aspirations.
While there’s excitement surrounding Manus, skepticism persists, particularly concerning its originality and technical robustness. Critics have dismissed it as merely a “glue factory” for existing AI models. Bradford Levy, assistant professor at the University of Chicago, stated, “From the demonstrations observed, it appears Manus does not always produce the expected outcomes on the initial attempt.” This critique is complemented by calls for users to be vigilant about data privacy, as the National Intelligence Law mandates cooperation with state intelligence agencies, raising concerns about information security.
Organizations instituting Manus technology have reported increased efficiency—between 15% to 25%—for tasks related to finance and human resources. Users can delegate complex workflows to Manus, allowing them to wake up to finished projects completed autonomously overnight. This facet emphasizes Manus as less of a competitor to human labor and more of a useful assistant allowing professionals to focus on higher-level strategic tasks.
The underlying philosophy advocated by Manus embraces the notion of “extend your capabilities,” as expressed by co-founder Xiao Hong. The name Manus draws inspiration from the Latin phrase Mens et Manus, meaning “mind and hand,” reflecting the ideology of merging knowledge and actionable outcomes.
Despite its promise, Manus is not without technical issues. Early testers encountered server overloads and error messages as the program processes extensive inquiries. One beta tester recounted, “Unfortunately, Manus AI failed after 50 minutes at step 18/20, showcasing the software’s limitations.” This prompted calls from users for improvements in system reliability and speed, especially when dealing with extensive data requests.
Moving forward, Butterfly Effect faces decisions on whether to expand Manus across consumer markets or focus on bespoke solutions for enterprise clients. The dynamics of the competitive marketplace will demand adaptability as the AI sector evolves rapidly. Manus’s scalable SaaS model could serve them well as they integrate user feedback for continued refinement.
Overall, Manus exemplifies the exciting potential of AI and signals China’s increasing focus on practical, impactful AI solutions rather than solely pursuing theoretical advancements. The future of AI, as illustrated by Manus, is about usability and practical applications—a paradigm shift poised to redefine the industry's standard.