After DeepSeek, China’s New AI Agent “Manus” is Automating Everything Even More Powerful?
By March 6th, 2025, a Chinese company introduced an AI News about its AI Agent, named Manus AI. One kind of powerful new AI tool, creating change not only in China but also echoing across the globe, raising hopes that it could replicate the success of DeepSeek, which earlier this year rattled the global tech industry with its state-of-the-art chatbot.
Manus AI originates from the Latin phrase “Mens et Manus” (mind and hand), echoing the artificial intelligence solutions‘ fundamental philosophy of applying knowledge to practical tasks rather than merely responding to queries.
According to its website, Manus, an AI agent generally considered more advanced than a chatbot, can do everything from analyzing the stock market to creating a personalized travel handbook for a trip with simple instructions from users.
Due to its incredible and Advanced Work Environment, this agent is a step above an AI chatbot. While a chatbot can engage in limited conversations with a user, this AI Chatbot is capable of functioning as a virtual employee, performing activities with little to no user interaction.
Manus AI claimed that its Artificial intelligence Solutions (Chinese AI Manus) are better than those of its competing OpenAI’s AI agent, DeepResearch, which met its expectations in fulfilling the ‘incredibly challenging’ expectations set for it.
On the other hand, according to statistics released by Data Camp, Manus AI has shown a remarkable performance on the GAIA benchmark, a consistent test for AI agent problem-solving capabilities. At the basic task level, Manus scored 86.5%, significantly outperforming OpenAI’s
Deep Research system at 74.3%. Manus scored 70.1% for quick tasks, slightly surpassing OpenAI’s 69.1%. Even at the most challenging level involving complex, multi-step reasoning, Manus led with 57.7% compared to OpenAI’s 47.6%.
Chinese social media erupted when the AI news of the launch surfaced, as many people compared manus AI assistant to the vision of the creators of the DeepSeek chatbot, who stunned Silicon Valley by launching an AI chatbot that competes with OpenAI’s best products but at a much lower price.
Due to the ramifications of DeepSeek’s entrance, the agreement with Qwen has the potential to destabilize the market even more.
Founded by Beijing Butterfly Effect Technology Ltd Co, Manus AI has headquarters in Beijing and Wuhan. It has promoted its brand by doing dozens of tasks for free on X, which has won the allegiance of many users. However, the company revealed on X that the invite-only AI agent and the one with the ever-increasing website problems also malfunction.
What Did We Experience with Manus AI?
After AI News of Manus, our team checked it out as a highly usable and feature-rich application with or without programming skills. For two of the three tasks, it outperformed ChatGPT DeepResearch, but it was much slower. Manus appears best for detailed analytical work that requires a lot of open-ended internet research, but is narrowly defined. In other words, it’s more useful for the tasks a skilled human intern could complete in a working day.
However, it’s not all rainbows and butterflies. Frequent crashes and system instability is a common problem for Manus, and it has difficulty trying to process large parts of text. When my group and I were trying to make new requests, “tasks cannot be created because of high service load, we apologize for the inconvenience, please try again in a few minutes” popped up on my screen a few times, and Manus’s Computer would sometimes get stuck on a page for a long, long time.
Manus will definitely continue to be used for my various professional and personal tasks. Even though I don’t know if the DeepSeek comparisons are entirely accurate, this is more proof that Chinese Manus AI, not simply copying their Western colleagues. They are not just building on established model innovations, but are changing the narrative of how the world accepts autonomous Manus AI.
