Agents' Last Exam Unlicensed Full-Spectrum Pass Rate
Agents · 2026-06-03
Perfect-score pass rate on the 50-task unlicensed full-spectrum subset of Agents' Last Exam.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5.6 Sol | 30.0 |
| GPT-5.6 Luna | 28.0 |
| Kimi K3 | 28.0 |
| GPT-5.6 Terra | 26.0 |
| Claude Fable 5 | 24.0 |
| Opus 4.8 | 24.0 |
| GPT-5.5 | 22.7 |
| Opus 4.7 | 22.0 |
| Doubao Seed 2.1 Pro | 22.0 |
| GPT-5.4 | 19.3 |
| Composer 2.5 | 18.0 |
| GLM-5.2 | 18.0 |
| Gemini 3.1 Pro Preview | 12.0 |
| Qwen3.7-Max | 12.0 |
| DeepSeek-V4-Pro | 11.0 |
| GLM-5.1 | 10.0 |
| MiMo-V2.5 | 10.0 |
| Qwen3.6 Plus (2026-04-02) | 9.0 |
| Grok 4.3 | 8.0 |
| Kimi K2.6 | 7.0 |
| MiniMax M2.7 | 4.0 |