Chinese startup Monica has introduced Manus, an AI model capable of autonomously executing digital tasks, marking a major step forward.
Manus differentiates itself from traditional AI model chatbots by actively performing tasks rather than just generating ideas. The model operates on a virtual desktop, enabling it to conduct research, summarize findings, and compile reports autonomously. Unlike most AI systems that require human intervention, Manus can complete complex tasks such as screening resumes, extracting data, and organizing results in spreadsheets without user oversight.
Built on Anthropic’s Claude Sonnet model, Manus also supports multi-agent capabilities, allowing for more advanced reasoning and task execution. The system operates asynchronously in the cloud, meaning users can close their laptops and return to completed tasks later. According to benchmarks, Manus outperforms OpenAI’s Deep Research by 12%, positioning it as a competitive alternative in the AI landscape.
Manus represents a shift towards AI agents that bridge the gap between conception and execution. As Monica continues refining the model, future updates may introduce customizable AI agents and open-source tools, making advanced AI more accessible. While pricing and full availability remain unclear, Manus signals a move toward AI-powered autonomy in digital workflows.