Recently, a team from China officially released Manus, a general-purpose AI agent product. According to the team, Manus achieved SOTA (State-of-the-Art) results in the GAIA benchmark test, showing that its performance exceeds OpenAI's large model of the same level. From the introduction, compared with the current large model products, as a general-purpose agent, Manus is positioned as a powerful general-purpose assistant that not only provides ideas for users, but can put ideas into practice and truly solve problems.