Chinese Team Unveils AI Agent, Manus
On March 6th, coinciding with Apple’s new product launch, many people stayed up all night seeking invitation codes for Manus, the world’s first AI Agent product developed by Monica.im.
According to their team, Manus is a truly autonomous AI agent capable of handling various complex and dynamic tasks. Unlike traditional AI assistants, Manus not only provides suggestions or answers but also delivers complete task results.
Manus utilizes a multi-signature (multisig) system driven by multiple independent models. Later this year, the official plan is to open-source some of these models, particularly the ‘postering’ (inference) part of Manus.
Simultaneously gaining attention with Manus was a four-minute demonstration demo. In these cases, Manus autonomously completed the entire process from planning to execution, showcasing genuine Agent capabilities rather than simple assistant functions.
For example, the first task involved screening suitable candidates for a reinforcement learning algorithm engineer position from 15 resumes. Manus demonstrated qualities akin to a human intern, manually unzipping files and reviewing each resume page by page, while recording key information. The official team stated that what was shown represents only the tip of the iceberg of Manus’s capabilities.
When the AI Agent, through invoking a long chain of thought and tool utilization, ultimately outputs a complete and professional result, users began to marvel at the fact that AI can truly help humans accomplish tasks.
To ensure the reproducibility of results, Manus was evaluated using configurations identical to its official version.
According to the official website, in the GAIA benchmark test (which assesses the ability of general AI assistants to solve real-world problems), Manus achieved new state-of-the-art (SOTA) performance across all three difficulty levels.
The founder behind Manus AI, Xiao Hong, graduated from Huazhong University of Science and Technology with a degree in software engineering.After graduation, he embarked on a series of entrepreneurial ventures, founding Ye Ying Technology in 2015 and launching Yi Ban Assistant and Wei Ban Assistant, serving over 2 million B-end users and securing investments from Tencent, ZhenFund, and others.
He also developed Monica, an AI assistant product touted as ‘All-in-One,’ initially launched as a browser plugin. By integrating mainstream large models (such as Claude 3.5, DeepSeek, etc.), Monica provides functions such as chat, translation, and copywriting. Monica initially focused on overseas markets, reaching a user base of over one million and becoming a leading product in the AI plugin field. In February of this year, the Chinese version of Monica began internal testing and is currently available for free to domestic users. This version is built on the DeepSeek R1 and V3 models, possessing deep reasoning capabilities and supporting memory functions and real-time internet search.
Manus adheres to the technical philosophy of ‘less structure more intelligence,’ which differs somewhat from the mainstream. They believe that when the data is of sufficient quality, the model is powerful enough, the architecture is flexible enough, and the engineering is robust enough, capabilities such as computer use, deep research, and coding agents will naturally emerge, without needing to be designed as specific product features.
Currently, almost all technology companies in the AI sector are focusing on ‘AI agents.’ On March 6th, OpenAI announced that it will charge $20,000 (approximately 145,000 RMB) per month for doctor-level AI Agents, primarily targeting the high-end needs of enterprise users, especially in data-intensive industries such as finance, healthcare, and manufacturing.
SEE ALSO: Tesla FSD Officially Launched in China
