While there may not be a universally accepted definition of an AI agent, numerous startups are emerging with the goal of creating agentic tools designed to automate various online tasks. One such innovative company, Browser Use, has garnered significant attention from both developers and investors for its groundbreaking solution that enhances the readability of websites for AI agents. Recently, Browser Use announced it has successfully raised a substantial $17 million seed funding round, led by Felicis’ Astasia Myers, with participation from notable investors such as Paul Graham, A Capital, and Nexus Venture Partners. This funding round had not been previously disclosed.
Browser Use is part of Y Combinator's 2025 winter batch and has gained notoriety in recent months, particularly due to its association with the Chinese startup Butterfly Effect, which utilized Browser Use in its viral Manus tool. The collaboration has significantly elevated Browser Use's visibility and reputation in the tech community. Founded by Magnus Müller and Gregor Zunic last year through ETH Zurich’s Student Project House accelerator, Browser Use was born out of a shared vision to streamline web interactions.
Müller, who has extensive experience in developing web-scraping tools, met Zunic in 2024 while both were pursuing their master’s degrees in data science. The duo conceptualized the idea of merging web scraping techniques with data science principles to enable a browser to perform specific tasks autonomously. In just five weeks, they built a demo of Browser Use, which quickly gained traction and was subsequently made open-source, allowing more users to benefit from its capabilities.
The core functionality of Browser Use revolves around its ability to simplify website elements and buttons into a more digestible, text-like format that AI agents can easily comprehend. This transformation assists AI agents in understanding various options available on a website, enabling them to make decisions independently. Müller emphasized the limitations of traditional vision-based systems, stating, “A lot of agents rely on vision-based systems and try to navigate websites through screenshots, and in the process, things break.” By converting websites into a format that agents can understand, Browser Use allows for the repetition of tasks at a lower cost.
As the demand for AI solutions that facilitate smoother interactions with websites grows, Müller believes that Browser Use has the potential to become a fundamental layer in this evolving landscape. He highlighted that over 20 companies in the current Y Combinator Winter batch are utilizing Browser Use to fulfill their specific needs. Müller noted, “There are companies coming to us and saying, ‘What can we do to make it easier for agents to navigate our website?’” He pointed out that many websites, like LinkedIn, frequently alter their structures, causing agents to struggle with navigation.
According to Astasia Myers from Felicis, the firm has been actively exploring the AI agents sector for several years, and Browser Use presented an ideal opportunity to expand their investment portfolio in this area. Myers remarked that the company’s founding team and their open-source-first approach were pivotal factors in their decision to invest. She stated, “We think web AI agents are the next frontier that really helps with the end-to-end automation of human tasks. Web AI agents serve as a dynamic bridge between static pre-trained models that are primarily text-focused and the rapidly changing digital landscape.”
