Benchmark artifacts
Browse the benchmark artifacts.
Every evaluated task, grouped by category and tiered by claim. Open a category to see the task prompts and the generated artifacts from each tool.
System prompt provided with each task
- - Use ONLY HTML, CSS, and JavaScript (no frameworks or libraries).
- - Do NOT use web search or external tools.
ChatGPT
- Structural8 tasks
- Styling8 tasks
- Functional8 tasks
24 tasks
Claude
- Structural8 tasks
- Styling8 tasks
- Functional8 tasks
24 tasks
Firebase
- Structural8 tasks
- Styling8 tasks
- Functional8 tasks
24 tasks
Bolt
- Structural8 tasks
- Styling8 tasks
- Functional8 tasks
24 tasks
v0
- Structural8 tasks
- Styling8 tasks
- Functional8 tasks
24 tasks