Enhans, a South Korean startup specializing in vertical commerce AI agents, announced that its artificial intelligence model, ACT-1, secured third place in the global Online Mind2Web benchmark for web-based AI agents. The model ranked just behind global technology leaders OpenAI and Anthropic. The Online Mind2Web benchmark evaluates AI agents on their ability to perform complex tasks on the web by emulating human behavior. It covers 300 real-world tasks across 136 actively operating websites, assessing how effectively AI agents can navigate and interact with diverse web environments. ACT-1 is a behavior-based AI model designed specifically for web automation. It can execute sophisticated actions such as clicking buttons and navigating menus. Unlike conventional AI approaches that rely on screenshots and coordinate-based controls, ACT-1 uses a document object model (DOM)-based method to directly access and manipulate webpage structures. Enhans emphasized that ACT-1 achieved the No. 1 ranking in the commerce domain among global competitors, highlighting the potential of Korean-developed AI agents on the international stage. Building on this momentum, Enhans plans to evolve ACT-1 into a next-generation AI agent by integrating imitation learning techniques. This advancement aims to enable the AI to perform tasks traditionally limited to human users. Additionally, Enhans intends to expand its enterprise-level web automation solutions, aiming to differentiate itself from global AI firms and secure a unique competitive edge. “Being able to stand alongside global AI leaders like OpenAI and Anthropic is profoundly meaningful. We will continue to strengthen the development of our web AI agent technologies to lead the global market in vertical industries,” said Enhans CEO Lee Seung-hyun.

Frequently Asked Questions (FAQ)

Q: What is the Online Mind2Web benchmark? A: The Online Mind2Web benchmark is a global evaluation that assesses AI agents on their capability to perform complex tasks on the web by imitating human behavior. It examines their performance across a wide range of real-world tasks on operational websites. Q: How does Enhans' ACT-1 model differ from traditional AI approaches for web automation? A: Unlike conventional AI methods that rely on screenshots and coordinate-based controls, ACT-1 utilizes a Document Object Model (DOM)-based approach. This allows it to directly access and manipulate webpage structures, leading to more robust and sophisticated web automation. Q: What are Enhans' future plans for ACT-1? A: Enhans plans to further develop ACT-1 by integrating imitation learning techniques. This will enable the AI agent to perform tasks that are currently typically handled by human users, expanding its capabilities. Q: In which domain did ACT-1 achieve the number one ranking? A: ACT-1 achieved the No. 1 ranking in the commerce domain among global competitors in the Online Mind2Web benchmark. ##Crypto Market AI's Take This development from Enhans highlights the rapid advancement and increasing sophistication of AI agents, particularly in the realm of web automation. Our platform, Crypto Market AI, is at the forefront of integrating AI into financial markets, offering tools that leverage similar AI capabilities for market analysis, trading, and portfolio management. The success of ACT-1 in a competitive benchmark underscores the potential for AI agents to revolutionize various industries, including the dynamic cryptocurrency sector. As AI agents become more adept at understanding and interacting with complex digital environments, their application in automating trading strategies, identifying market trends, and enhancing user experience in financial platforms is set to grow significantly. ##More to Read:

Source: Originally published at The Korea Herald on July 29, 2025.

Korea's Enhans makes global top 3 in AI agent benchmark

Frequently Asked Questions (FAQ)