WebsiteBench
WebsiteBench is an Apache-2.0 benchmark from TIGER-AI-Lab that asks whether AI agents can rebuild a production website just by exploring it in a browser. Agents must reverse-engineer a full-stack copy that matches the original's appearance and behavior; the set has 70 websites and 46,916 hidden test cases, downloadable with the websitebench PyPI package and run in Docker.
Updated
✨ Try AgentScout's free AI image generator →Good for: browsing the web, research, evaluating agents, testing apps
Use cases
About WebsiteBench
WebsiteBench is listed in our evals & observability category and is aimed at builders. WebsiteBench is an Apache-2.0 benchmark from TIGER-AI-Lab that asks whether AI agents can rebuild a production website just by exploring it in a browser. Agents must reverse-engineer a full-stack copy that matches the original's appearance and behavior; the set has 70 websites and 46,916 hidden test cases, downloadable with the websitebench PyPI package and run in Docker. Pricing, as the maker states it: Open source. Maturity: Active. We last checked the official link on 2026-11-18.
WebsiteBench at a glance
- Category: Evals & observability
- Built for: Developers and teams building with AI agents
- Pricing: Open source (as the maker states it)
- GitHub: TIGER-AI-Lab/WebsiteBench · 528 stars (as of Nov 18, 2026)
- Official site: website-bench.com
- Listed on AgentScout: Nov 18, 2026
WebsiteBench FAQ
What is WebsiteBench?
WebsiteBench is an Apache-2.0 benchmark from TIGER-AI-Lab that asks whether AI agents can rebuild a production website just by exploring it in a browser. Agents must reverse-engineer a full-stack copy that matches the original's appearance and behavior; the set has 70 websites and 46,916 hidden test cases, downloadable with the websitebench PyPI package and run in Docker.
How much does WebsiteBench cost?
WebsiteBench's pricing, as the maker states it: Open source. Open-source tools are free to use, though you may still pay for model API usage or a hosted plan. Check the official site for current plans.
Where can I find WebsiteBench's source code?
On GitHub at TIGER-AI-Lab/WebsiteBench, which had 528 stars as of Nov 18, 2026.
What are alternatives to WebsiteBench?
Similar tools include SWE-bench, Tracely, Opik, Basalt, CoreWeave Forge and OSWorld.
Facts from our link-checked listing, the GitHub repository and the official site. See something wrong? Suggest an edit.
WebsiteBench ratings and reviews
Be the first to rate WebsiteBench: tap 1 to 5 stars below. A written review is optional.
WebsiteBench alternatives
Other evals & observability tools in our directory. See all evals & observability tools →
SWE-bench
Benchmark that tests whether language models and coding agents can resolve real GitHub issues, with multimodal and verified variants.
Tracely
Tracely is open-source, trace-native CI/CD for AI agents: it grades every production trace with evaluators, clusters failures into issues, freezes bad runs into replayable test cases and blocks pull requests that would reintroduce them.
Opik
Opik is Comet's Apache-2.0 platform for LLM observability and evaluation: it traces LLM calls, RAG pipelines and agent runs, scores them with built-in and custom metrics (including LLM-as-judge), runs experiments on datasets, manages prompts and monitors production traffic with dashboards. SDKs and integrations cover Python, TypeScript, OpenAI, LangChain, LlamaIndex and other frameworks. The full platform is free to self-host, or you can use Comet's hosted cloud.
Basalt
Captures customer behavior so your AI agent can learn and improve from real interactions.
CoreWeave Forge
One product to run, observe, evaluate and post-train agents (W&B, OpenPipe, sandboxes).
OSWorld
Benchmark for multimodal computer-use agents on open-ended tasks in real desktop environments.