ToolBench - Tabletop — leaderboard
Metric: Task Score. Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4 | 81 |
| 2 | text-davinci-003 | 66.7 |
| 3 | CodeLlama-13B-Instruct-hf | 57.62 |
| 4 | CodeLlama-34B-hf | 51.32 |
| 5 | CodeLlama-34B-Instruct-hf | 47.46 |
| 6 | Llama 2 70B | 45.4 |
| 7 | CodeLlama-13B-Python-hf | 41.53 |
| 8 | CodeLlama-13B-hf | 40.95 |
| 9 | LLaMA-30B | 34.3 |
| 10 | CodeLlama-34B-Python-hf | 33.33 |
| 11 | GPT-3.5 Turbo | 33.3 |
| 12 | LLaMA-65B | 30.5 |
| 13 | Llama 2 13B | 23.81 |
| 14 | CodeLlama-7B-Python-hf | 22.86 |
| 15 | starcoder | 21.9 |
Interactive version: theaggregate.ai/benchmark?slug=toolbench-tabletop · How the rankings work · Data refreshed daily, snapshot 2026-07-22.