The Chinese AI Agent market is experiencing a rapid evolution, with a new contender named Hermes quickly emerging in just two months to challenge the prominence of OpenClaw (popularly known as "Lobster"). This swift rise has garnered significant industry attention, prompting an examination of Hermes' unique appeal despite its seemingly similar functionalities to its predecessor.
OpenClaw had previously been a phenomenal product in China's AI Agent landscape. Its ability to empower users to build local AI assistants on Windows or Mac swiftly ignited genuine demand for AI Agents. This momentum was so strong that major tech giants like Tencent, Alibaba, and Baidu responded within 48 hours, integrating OpenClaw services into their ecosystems. For instance, Tencent developed local AI assistants based on OpenClaw and provided integration solutions with WeChat enterprise intelligent robots, even launching a zero-deployment version called DuClaw to lower user entry barriers.
However, OpenClaw's rapid ascent was not without its challenges. On the security front, the 360 team identified a critical 0day vulnerability in the OpenClaw Gateway WebSocket, involving unauthenticated upgrades, which was officially confirmed. Subsequently, the National Internet Emergency Center issued a risk advisory for OpenClaw's security applications. From a user experience perspective, some users reported issues such as insufficient scraping capabilities and complexities in crafting precise instructions, leading to a high trial-and-error cost for "raising shrimp" (i.e., deploying AI Agents), requiring extensive time for fine-tuning and optimization.
Against this backdrop, Hermes' rapid emergence is particularly noteworthy. While specific technical details or core advantages of Hermes are not extensively detailed in the original information, its ability to quickly become a new focal point amidst OpenClaw's popularity suggests significant innovations. These could include advancements in deployment convenience, enhanced security features, performance optimization, or improved user experience (e.g., more intelligent skill management, intuitive instruction interaction). This not only reflects the rapid iteration within China's AI Agent market but also indicates that industry competition is shifting from basic functional implementation to deeper considerations of user value and security.
Hermes' swift succession represents more than just a product change; it is a microcosm of the increasingly mature and diversified AI Agent ecosystem in China. It compels the industry to consider how to balance innovation with security and enhance practical user experiences in the wave of AI Agent popularization, which will be crucial for future competition.