Step 2 of 4 · ProjectsBenchmarks
HSCodeComp
Expert-level benchmark for HS code classification agents on e-commerce products, from Alibaba's MarcoPolo.
What it does
A benchmark of 632 expert-annotated e-commerce products, each labelled with a 10-digit HS code, for testing whether search agents can apply hierarchical tariff rules; an evaluation script reports exact-match accuracy at 2 to 10 digits. The data is a CSV plus a Python script needing pandas and an OpenAI-compatible key. Apache 2.0; the card cannot rule out copyright issues in the product data.
Where it fits on the map
Similar projects
Of the same kind, for a shared jurisdiction or tax area. Not a ranking.
- Legora transfer pricing caseEvaluation case for AI: draft a transfer pricing Master File from case documents, graded on 175 criteria.
- PolicyBenchBenchmark of how well frontier AI models estimate household tax and benefit outcomes without tools.
- UK Trade Tariff AI search evaluation suiteEvaluation suite for the UK tariff team's AI classification and search experiments on the Online Trade Tariff.Demo
Maintain this project?
Link to this page from your README, so readers find the facts, the notes and where it fits. The badge is a small SVG served from this site.
[](https://tax-oss.com/projects/hscodecomp/)