HSCodeComp - 2-digit (Vision): leaderboard
HSCodeComp 2-digit chapter accuracy for the vision arm, where models additionally receive the product images alongside the text attributes.
Metric: Exact Match Accuracy (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 82.28 | #152 |
| 2 | GPT-5 | 82.12 | #98 |
| 3 | Claude Sonnet 4 | 78.8 | #199 |
| 4 | GPT-4o | 78.01 | #343 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hscodecomp-2-digit-vision · How It Works · Data refreshed daily, snapshot 2026-10-20.