Gorilla Benchmark API Bench: leaderboard
1,645 ML-model API calls (925 HuggingFace, 626 TensorFlow Hub, 94 TorchHub) paired with 10 synthetic instructions each, from UC Berkeley (2023); generated calls scored by AST sub-tree matching.
Metric: Score (%). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.1 405B Instruct | 35.3 | #457 |
| 2 | Llama 3.1 70B Instruct | 29.7 | #561 |
| 3 | Llama 3 70B Instruct | 14.7 | #634 |
| 4 | Llama 3.1 8B Instruct | 8.2 | #1035 |
| 5 | Llama 3 8B Instruct | 1.7 | #1129 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=gorilla-benchmark-api-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.