Chongqing - The tech industry is abuzz with excitement following the debut of Manus, a newly launched AI Agent hailed as the world’s first general AI agent. Within 24 hours of its release, Manus became the most talked-about product in China’s AI community. 

Unlike conventional AI assistants, Manus doesn’t just generate responses—it takes action. Users can simply input a task, and Manus autonomously delivers a near-ready planning report, financial analysis, travel itinerary, or PowerPoint presentation with minimal user intervention. The hype surrounding this breakthrough has sent demand soaring, with early-access invitations reportedly selling for as much as 100,000 yuan (about 14,000 U.S. dollars). But is Manus truly a revolutionary innovation, or just an overhyped trend?

Manus is a newly launched AI Agent hailed as the world’s first general AI agent.

More than just a chatbot

Manus has set itself apart with a hands-on approach to task execution. Its official description states, “Manus transforms ideas into actions, going beyond thinking to focus on tangible outcomes.” The AI agent can autonomously break down complex tasks into subtasks, execute them step by step, and produce structured results.

For instance, in resume screening, Manus doesn’t just analyze resumes—it unpacks compressed files, scans multiple documents, extracts key details, ranks candidates, and categorizes them based on work experience, all without requiring additional prompts from the user.

In real estate research, a user can instruct Manus to find a safe, family-friendly neighborhood in New York and identify suitable properties. Manus methodically gathers crime rate statistics, school rankings, and financial data, even writing a Python script to calculate affordability. It then compiles all findings into a comprehensive report, streamlining a process typically requiring multiple tools and manual effort.

Similarly, in stock market analysis, Manus retrieves historical stock data from financial platforms, cross-verifies sources to ensure accuracy, performs statistical analysis with Python, and presents results through data visualizations and expert-level financial reports.

Unlike conventional AI assistants, Manus does not simply provide suggestions—it takes action, calling external tools, executing code, and interacting with various applications. This distinguishes it from AI models like DeepSeek, which primarily focus on deep thinking and conversation. In short, Manus isn’t just about answering questions—it delivers completed tasks.

The brains behind Manus

The rise of Manus has also drawn attention to its creators. Monica. im, the company behind it, is led by a team of three co-founders with backgrounds in AI and entrepreneurship.

Xiao Hong, a Huazhong University of Science and Technology graduate, founded Nightingales Technology in 2015, creating enterprise tools used by over 2 million businesses, later launching the AI assistant Monica, which amassed more than 4 million users globally.

Another key figure is Ji Yichao, a Beijing-based AI expert known for developing Magi, China’s largest general knowledge graph, and the Steiner open-source AI model, which replicated OpenAI’s early architecture.

The third co-founder, Zhang Tao, previously worked at ByteDance, contributing to Manus’ multi-agent system design and emphasizing “results-driven AI” rather than traditional chatbot-style responses.

Marketing tactics spark controversy

Despite Manus’ impressive capabilities, its marketing strategy has raised eyebrows. The product launched with an invitation-only system, and early access is restricted to users who register with an international payment method, sparking accusations of “artificial scarcity” and an “export-first, domestic-later” approach. This exclusivity drove demand so high that invitation codes were resold for up to 100,000 yuan (about 14,000 U.S. dollars) on secondary markets.

Responding to criticism, co-founder Zhang Tao clarified that the limited access was due to server capacity constraints, not a deliberate scarcity tactic.

Another debate surrounds Manus’ technological foundation. Unlike AI models that develop their own large-scale language models (LLMs), Manus relies on Claude, an AI model developed by Anthropic. This has led to accusations that Manus is merely “wrapping” third-party AI models rather than innovating independently. However, industry experts argue that this critique misses the point.

Experts defend Manus

Huang Shan, co-founder of Chongqing Sringcloud Digital & Technologies Co., Ltd. and an AI researcher at Shanghai Jiao Tong University Chongqing Institute of Artificial Intelligence, believes Manus represents an important step toward the AI Agent era.

“Critics calling Manus a ‘wrapper’ fail to see the bigger picture,” he stated. “What matters is not whether it built its own LLM, but whether it is pioneering a new direction for AI development.”

Manus’ launch underscores the growing demand for AI agents capable of autonomous decision-making and task execution. Some experts predict that 2025 will mark the commercial breakthrough of AI Agents, much like DeepSeek revolutionized AI models.

As AI continues to evolve, the industry eagerly awaits whether Manus will lead the shift toward fully autonomous AI Agents—or if another competitor will rise to claim the spotlight.

(Li Zuoting, a reporter from Shangyou News, contributed to this report's Chinese version.)