Introducing the Studio, our platform for developers to run automations at scale, and Runner H, the most advanced agent to date — designed to navigate web interfaces through pixel-level interpretation and semantic understanding. You can now turn your instructions into action with human-like precision.

Runner H is powered by a specialized LLM highly optimized for function calling and coding (2B parameters). It also integrates a specialized VLM (3B parameters) to perceive, understand, and interact with graphical and text elements. Runner H delivers the state-of-the-art on the public WebVoyager benchmark.

Our goal extends beyond optimizing AI agent performance metrics—we're expanding their operational scope and capabilities. We've engineered Runner H to handle any web interaction with enterprise-grade reliability, advancing toward one singular mission: automating the web.

With the beta release of Runner H, we are breaking through the limitations of plateauing scaling laws, taking a step toward artificial super intelligence (ASI). With Runner H, we’re not just introducing a product; we’re opening a new chapter in AI.

Ready to experience #RunnerH firsthand and see what it can do for you? Sign up for the private beta waitlist and learn more about our vision at

Nice, glad to see the space growing! We have build our own OSS version of computer use called

I hate these deceiving and misleading BS graphs:

Existing automations will now become 100x times more powerful with AI 💪

This looks very promising.

Can it manipulate the canvas? Like draw something in Figma for example then export it as SVG (canvas and html interactions)

So excited to check this out!