On March 5, 2025, the world's first general AI agent application ManusAI officially debuted, sparking a dual wave in Technology and Capital Markets.

This product, developed by a Chinese team, claims to automatically execute complex tasks such as deep research, data analysis, and contract review, thanks to its multi-model driven architecture and engineering orchestration capabilities. HTSC's Research Reports define it as a "key breakthrough in the Action phase of agent applications," predicting that 2025 could become the year of large-scale AI Agent applications.

The day after the product launch, the price of the Manus internal testing invitation code on second-hand trading platforms skyrocketed from 999 yuan to 0.1 million yuan, and a gray service for writing application emails emerged on social platforms.

However, behind the market frenzy, the technical test results showed significant discrepancies: some testers confirmed that it could run programs in the cloud and deliver complete website interactive results, with code writing and visual generation cases demonstrated by tech bloggers attracting widespread attention; yet in practical tests within the Education field, users uploading second-grade Chinese learning plans experienced interface chaos during the generation process, and a steam installation package download test ended with invalid links. Multiple programmer users pointed out the system's complex task response delay issue. This duality in technical performance reflects the reality challenges for AI Agents transitioning from the laboratory to commercialization.

Starting March 6, several tokens named 'Manus' emerged in the cryptocurrency market, revealing speculators' attempts to profit from this technology hype. Insiders close to the Manus team urgently refuted these claims, emphasizing that the token issuance has no connection to the official team.

The next day, the freezing incident of the official Manus X account escalated the controversy, with co-founder Ji Yichao stating that the account freeze may be related to cryptocurrency fraud impersonation, announcing the initiation of legal proceedings to combat infringement, and reiterating their stance of "never having ventured into the blockchain field."

In the face of the ongoing market chaos, the Manus team has been vocal with frequent updates. Partner Zhang Tao disclosed that the server resources originally configured according to industry demo standards have become overwhelmed, and the technical team is urgently expanding to cope with the unexpectedly high access pressure. Under the principle of "zero market budget," the team committed to optimizing the internal testing process, prioritizing the existing user experience, and gradually and orderly releasing invitation qualifications. Ji Yichao emphasized that an official verification channel would be established and revealed that a more powerful task planning module is being developed to address current tool utilization bottlenecks.

From the exorbitant prices of invitation codes in the second-hand market to the account freezing turmoil on social platforms, the explosive trajectory of Manus reveals the deep contradictions in the AI Agent track— the gap between the Capital Markets' urgent expectations for technology implementation and the maturity of the technology has not yet been bridged.

Industry observers point out that Manus's engineering breakthroughs are milestone achievements, but its stability in complex scenarios, the identity verification mechanism under a compliance framework, and the response capability to sudden public opinion crises will all be key variables affecting the industry's direction. As tech giants like Google and Microsoft accelerate their layout of the Asia Vets ecosystem, the AI Agent storm initiated by Manus may be reshaping the competitive landscape of the Global AI industry.

Editor/danial