ACEBench: leaderboard
2,000 tool-calling test cases in parallel English and Chinese over 4,538 APIs, split into Normal, Special (ambiguous instructions) and multi-turn Agent sets (Huawei Noah's Ark Lab, 2025); accuracy.
Metric: Overall (imported). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 29 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o (2024-11-20) | 89.6 | #381 |
| 2 | GPT-4 Turbo | 88.6 | #404 |
| 3 | Qwen Max | 81.7 | #382 |
| 4 | O1 Preview | 80.6 | #193 |
| 5 | DeepSeek V3 Chat | 78.5 | #447 |
| 6 | GPT-4o Mini (2024-07-18) | 76 | #558 |
| 7 | Claude 3.5 Sonnet (20241022) | 75.6 | #291 |
| 8 | Gemini 1.5 Pro | 72.8 | #445 |
| 9 | O1 Mini | 72.2 | #323 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=acebench · How It Works · Data refreshed daily, snapshot 2026-10-20.