TIGER-AI-Lab releases WebsiteBench benchmark
Understand this story
TIGER-AI-Lab has released WebsiteBench, a benchmark that requires AI agents to explore production-grade websites through a browser and then reverse-engineer a full-stack version whose appearance and behavior match as closely as possible. The benchmark provides local-tasks/ task packages, which can be run with the harbor run command together with agents such as codex and a specified model, with results written to jobs/. The code is open-sourced under the Apache-2.0 license, while task data and reference content follow their respective licenses.
AI overview based on source reports · Updated 17 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIWebsiteBench: Can AI Agents Rebuild Websites Through Browser Exploration?
WebsiteBench is a benchmark in which agents explore a production-level website through a browser and must reverse-engineer a full-stack version matching its appearance and behavior as closely as possible. The repository provides quickstart instructions using uv, Docker, and the harbor run command to evaluate single tasks from local-tasks/ or all tasks, with results written to jobs/. Code is licensed under Apache-2.0, while task data and reference content retain their applicable licenses.
Activity for this story
- Current comparable activity
- 6
- Peak comparable activity
- 10Nov 5 21:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.