A
AutoMedBench Lite
Efficiency Frontier
Leaderboard
OpenRouter
Overall vs.
Cost per Task
Overall vs.
Time per Task
Overall vs.
Tokens per Task
Overall vs.
Turns per Task
7-track average
Overall score vs. Cost per Task
28 of 28 models
</>
Open Source Only
Most attractive quadrant
Pareto line
☾ Dark mode