EComStage - Solution Decision: leaderboard
Evaluates text-based e-commerce assistance across seven perception, planning and action tasks. The overall score and task results share the main-table evaluation protocol. This series reports Solution Decision using Task Score (%). Source: the original paper result table.
Metric: Task Score (%). Source: arxiv.org. Saturation forecast: Around February 2027. 33 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 30B A3B Instruct | 94.25 | #487 |
| 2 | Qwen 3 235B A22B Instruct | 89.94 | #336 |
| 3 | Qwen 3 235B A22B | 87.47 | #323 |
| 4 | Gemini 2.5 Pro | 87.06 | #159 |
| 5 | Qwen 2.5 72B Instruct | 87.06 | #469 |
| 6 | Qwen 3 8B | 85.01 | #678 |
| 7 | Qwen 3 4B Instruct | 84.8 | #661 |
| 8 | Qwen 3 30B A3B | 81.72 | #496 |
| 9 | InternVL3-14B | 81.11 | #506 |
| 10 | Qwen 2.5 14B Instruct | 80.9 | #654 |
| 11 | GPT-4o | 79.47 | #357 |
| 12 | Qwen 2.5 1.5B Instruct | 78.85 | #1393 |
| 13 | GPT-OSS-120B | 78.23 | #346 |
| 14 | Qwen 3 4B | 78.23 | #834 |
| 15 | Qwen 2.5 3B Instruct | 77.62 | #1159 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ecomstage-solution-decision · How It Works · Data refreshed daily, snapshot 2026-11-05.