The bitter lesson of browser agents

As models get better, the browser harness has to change. A lot.

When we launched Browser Use in November 2024, GPT-4o wasn't trained for computer use. Built on next-token prediction, it didn't reliably understand what it could do on a page. We had to spell it out.

As models got better at coding, we moved from predefined state and actions to code: first the model wrote its actions, then its observations. Now it does both through raw CDP, inside an existing agent harness.

We defined the state: here's what's on the page, here's what's clickable. Then the action space: click, type, scroll. Every interaction had to fit those rules.

The model picked an action, we ran it and sent back the next state. But browser automation is a game of edge cases. Every exception needed another fix.