It would be really cool to have an official 3D-print engineering benchmark/leaderboard like this

Someone prompted different LLMs to generate CAD code for a bridge under fixed constraints (2-foot span, under 500g filament, 18-hour print limit), printed them, and load-tested them to failure. The results were quite varied: some models couldn't even design parts that fit together, while the winner held over 100 lbs. Most benchmarks don't really capture physical intuition, spatial reasoning, and functional code generation at the same time. I would love a standardized benchmark and leaderboard for this.