Bench-bench v0.2, a public benchmark created by an independent developer, measures AI models' performance on long-horizon planning tasks, humorously framed as 'bench press' capability. The project aims to test whether AI models can handle sustained, multi-step reasoning, addressing a gap in current evaluation methods.
Bench-bench v0.2 · public run
How much can an AI model bench press?
I keep hearing labs talk about training, lift, and moving weights, but can their models even lift*? Bench-bench is my attempt to test that.
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.