China has introduced another powerful artificial intelligence model, Manus, marking a major step in the country’s accelerating AI race. Developed by the Chinese startup Monica, Manus arrives just weeks after DeepSeek’s launch and is already being compared to top AI models from OpenAI, Google, and Anthropic. Unlike conventional AI systems that require human input at multiple stages, Manus is designed to function independently, handling complex tasks without supervision.

Unlike standard AI chatbots that respond to queries, Manus is built to take direct action. It can design websites, create travel itineraries, analyze stock trends, and perform various other tasks based on a single prompt. For instance, if a user asks it to generate a report on climate change, Manus will conduct research, compile relevant data, create charts, and format the entire document into a structured report—all without needing further instructions.

Launched on March 6, Manus has already gained global recognition. According to Monica, it surpasses OpenAI’s DeepResearch in performance on the GAIA benchmark, an important AI evaluation metric. A demonstration video released by the company showcases Manus seamlessly interacting with the internet, retrieving information, and executing tasks in real time. It can browse websites, take screenshots, track online activity, and generate fully formatted reports, spreadsheets, and presentations. This capability has positioned it as a potential breakthrough in AI-driven automation.

One of Manus’s standout features is its ability to operate independently in the cloud. Even if a user disconnects, the AI continues working in the background and updates them when the task is complete. Unlike many AI models that generate static responses, Manus actively browses the web, interacts with online platforms, and displays its workflow as it gathers and processes data. Over time, it learns from user interactions, refining its output to match preferences and improve accuracy.

The AI model also has extensive integration capabilities. It can interact with platforms like X (formerly Twitter) and Telegram, process information from multiple sources, and even manage several screens simultaneously. It doesn’t just generate text-based results but produces interactive presentations, structured reports, and data visualizations.

Manus is currently available through an invitation-only web preview, with Monica yet to announce a public release date. However, the company has suggested that an official rollout could happen soon and has hinted at plans to open-source the model in the coming months. This decision could pave the way for developers to customize and expand its capabilities, potentially accelerating AI innovation and adoption worldwide.

As artificial intelligence continues to evolve, AI-powered browsers and autonomous agents like Manus are becoming a key focus for technology companies. The ability to conduct deep web research, automate complex workflows, and generate structured results could redefine productivity and decision-making processes. The growing excitement around Manus reflects the rapid acceleration of AI innovation, particularly in China. As competition intensifies among leading AI developers such as OpenAI, Google, and Monica, the world may be on the cusp of a new era where AI doesn’t just assist users but independently completes tasks from start to finish.

Disclaimer: This image is taken from X/@ChengduHitech.