Shopping Reasoning Bench: leaderboard
Metric: Weighted pass rate (%) over all 525 expert-authored shopping missions scored by importance-weighted binary rubrics written by retail domain experts (weighted pass rate); each model answers with its native web search tool at default inference settings and a Claude Sonnet 4.5 judge (temperature 0) grades every rubric; mission-weighted mean of the single- and multi-turn subsets; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 77.2 |
| 2 | Claude Opus 4.7 | 77 |
| 3 | Gemini 3 Flash | 75.5 |
| 4 | Gemini 3.1 Flash Lite | 72.4 |
| 5 | GPT-5.4 | 70.2 |
| 6 | Claude Sonnet 4.5 | 68.6 |
| 7 | GPT-5.4 Mini | 63.9 |
| 8 | GPT-5.4 Nano | 63.4 |
| 9 | Claude Haiku 4.5 | 57.4 |
Interactive version: theaggregate.ai/benchmark?slug=shopping-reasoning-bench · How It Works · Data refreshed daily, snapshot 2026-09-29.