Introduction

Holo is H Company’s Action Vision-Language Model (VLM) and Surfer-H is the agent that runs it. Together they interact with web interfaces the way people do: setting a goal, making decisions, reassessing where needed, and completing everyday tasks. Holo and Surfer-H can do everything from booking flights to searching for recipes online. The Surfer-H-CLI is the command-line interface for running and controlling the Surfer-H agent, allowing you to define tasks, connect to Holo models, and execute web interactions directly from your terminal.

​

Benefits

Holo with Surfer-H provides these benefits:

• Human-like interaction: Works with standard web UIs without needing special integrations.
• Efficiency: A range of models and model sizes that balance performance and cost.
• Goal-driven autonomy: Plans, executes, and re-evaluates actions to complete tasks.
• Open source and extensible: Transparent and customizable.
• Cross-platform: Runs locally, with Docker, or in the cloud (AWS SageMaker).
• Benchmarks: State-of-the-art performance on WebVoyager and several UI localization benchmarks.

​

Example use cases

Common Surfer-H use cases: