ToolBench - The Cat API — leaderboard
Metric: Task Score. Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | text-davinci-003 | 98 |
| 2 | GPT-4 | 96 |
| 3 | CodeLlama-7B-Instruct-hf | 94 |
| 4 | CodeLlama-13B-Python-hf | 92.35 |
| 5 | GPT-3.5 Turbo | 92 |
| 6 | CodeLlama-7B-hf | 92 |
| 7 | CodeLlama-13B-Instruct-hf | 89.58 |
| 8 | CodeLlama-34B-Python-hf | 88.42 |
| 9 | CodeLlama-7B-Python-hf | 88 |
| 10 | CodeLlama-13B-hf | 86.83 |
| 11 | CodeLlama-34B-Instruct-hf | 86 |
| 12 | CodeLlama-34B-hf | 85 |
| 13 | Llama 2 70B | 84.39 |
| 14 | LLaMA-30B | 84 |
| 15 | starcoder | 84 |
Interactive version: theaggregate.ai/benchmark?slug=toolbench-the-cat-api · How the rankings work · Data refreshed daily, snapshot 2026-07-22.