CoPE Reference - WebShop Success: leaderboard
Table 3 untrained agentic-prompting reference cohort; same simulated task split and prompt family; RL methods and Reflexion omitted because they change the system; source ScienceWorld raw reward may be negative and is not used as success. Metric: Success rate (%). Higher is better. Source: https://arxiv.org/html/2610.20597v1
Metric: Success rate (%). Source: arxiv.org. Saturation forecast: Around January 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 56.5 | #32 |
| 2 | GPT-5 Mini | 7 | #184 |
| 3 | Gemini 2.5 Flash Lite | 6.5 | #418 |
| 4 | Qwen 2.5 7B Instruct | 2.5 | #854 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=cope-reference-webshop-success · How It Works · Data refreshed daily, snapshot 2026-10-16.