ToolBench - WebShop Short: leaderboard
ToolBench subtask measuring shorter WebShop shopping-agent performance.
Metric: Task Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 44 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | LLaMA-65B | 41.2 | #1303 |
| 2 | Llama 2 13B | 31.67 | #1436 |
| 3 | LLaMA-30B | 30.6 | #1423 |
| 4 | Llama 2 70B | 30.45 | #1201 |
| 5 | LLaMA-13B | 27.6 | #1516 |
| 6 | CodeLlama-34B-Instruct-hf | 25.99 | #1381 |
| 7 | CodeLlama-13B-Instruct-hf | 9.7 | #1532 |
| 8 | Llama 2 7B | 6.92 | #1520 |
| 9 | CodeLlama-34B-Python-hf | 6.47 | #1552 |
| 10 | CodeLlama-34B-hf | 5.53 | #1515 |
| 11 | pythia-1.4B | 5.2 | #1626 |
| 12 | CodeLlama-13B-Python-hf | 2.4 | #1623 |
| 13 | GPT-3.5 Turbo | 1.8 | #865 |
| 14 | CodeLlama-7B-Python-hf | 1.58 | #1608 |
| 15 | GPT-NeoXT-Chat-Base-20B | 0.7 | #1586 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=toolbench-webshop-short · How It Works · Data refreshed daily, snapshot 2026-10-20.