Browser Use, an American startup aiming to transform how AI agents interact with the web, recently announced a $17 million seed funding round led by Felicis Ventures, with participation from A Capital, Nexus Ventures, Y Combinator, Paul Graham, Liquid2, SV Angel, and Pioneer Fund. 

Building the Interface Between LLMs and the Web 

Browser Use was founded in 2024 by Magnus Müller and Gregor Žunič, two former data science students from ETH Zurich. Their project began as a simple weekend experiment to test whether language models (LLMs) could navigate the web like humans. In just four days, they developed an initial prototype which they then launched on the Hacker News platform. The enthusiasm was immediate, confirming their intuition that the future of AI-driven web automation was closer than many thought. A few weeks later, the first demo was ready.
The two co-founders assert:
"The internet is the largest source of unstructured data in the world, but interacting with it still requires human actions: clicking buttons, filling forms, manually navigating websites. With the rise of LLMs and autonomous agents, this reality is changing. We are building the infrastructure that allows AI to interact with the web as naturally as a human."
Most existing automation solutions rely on vision-based methods, attempting to mimic human perception of web pages. Sensitive to visual variations (color changes, element positions...), they are, according to the two men, "slow, costly, and unreliable".  They adopted a radically different strategy: their tool converts web interfaces into structured text, allowing language models to interact with sites more predictably. This approach enables more precise interaction with user interface elements (buttons, forms, menus) while ensuring faster and more cost-effective execution than image analysis-based solutions.
Unlike most of its competitors, Browser Use Cloud can be used with different LLMs. Its Pro version is offered by the startup at $30/month, making it a more flexible and less expensive open-source alternative than OpenAI's Operator.

Rapid Traction and Diverse Use Cases

In just a few months, Browser Use has seen rapid growth. Its open-source project, actively contributed to by a community of over 15,000 developers, has garnered more than 48,400 stars on GitHub. Its tools cover various use cases, including automated login and web navigation, large-scale data extraction, quality assurance testing, and CRM integrations.
With this funding, Browser Use aims to accelerate the development of its infrastructure and plans to recruit top engineers for this purpose.
Magnus Müller comments:
"We firmly believe that the interaction between AI and the web will undergo a major transformation in the coming years. In a few years, we think AI-automated interactions will surpass those performed by humans."

Translated from Browser Use lève 17 millions de dollars pour accélérer le développement de son agent Web concurrent d'Operator

To better understand

What is the technology of converting web interfaces into structured text, and why is it important for Browser Use?

The technology of converting web interfaces into structured text allows language models to treat websites like textual databases, improving accuracy and reliability compared to visual methods. This is crucial for Browser Use as it reduces reliance on human perception and visual variations, making automation faster and more economical.

How does Browser Use's fundraising by various investors influence its development and innovation capabilities?

The diversity of Browser Use's investors, such as Felicis Ventures and Y Combinator, provides not only financial resources but also a strategic network of advice and partnerships. This accelerates its technological development and innovation while strengthening its position in the web automation field.