Benchmark
EvalCart
See whether a shopping agent can fill a grocery cart that fits a shopper’s budget and tastes while limiting waste.
Explore the original evidence
Explore the original evaluation, its methods, and reported findings.
Open evalCoverage
The creator, BSOhealth, described a simulated store with roughly 2,000 grocery items.
Methodology
After checkout, a receipt grades the cart on dimensions including completeness, budget, taste, waste and efficiency. Synthetic brands limit tests that require a specific real-world product. At launch, the leaderboard contained only the creator’s trials and baselines.