Agents' Last Exam Full-Spectrum Score
Agents · 2026-06-03
Average partial-credit rubric score on the 55-task full-spectrum subset of Agents' Last Exam.
Top models (higher is better)
| Model | Score |
|---|---|
| Kimi K3 | 49.9 |
| GPT-5.6 Sol | 46.6 |
| GPT-5.6 Terra | 45.6 |
| Opus 4.8 | 43.1 |
| Claude Fable 5 | 42.0 |
| GPT-5.6 Luna | 41.9 |
| GPT-5.5 | 41.1 |
| Doubao Seed 2.1 Pro | 40.1 |
| Opus 4.7 | 37.6 |
| GLM-5.2 | 37.0 |
| GPT-5.4 | 34.3 |
| Composer 2.5 | 30.8 |
| Qwen3.7-Max | 27.2 |
| Gemini 3.1 Pro Preview | 26.4 |
| DeepSeek-V4-Pro | 23.8 |
| Qwen3.6 Plus (2026-04-02) | 22.9 |
| GLM-5.1 | 21.7 |
| MiMo-V2.5 | 20.8 |
| Kimi K2.6 | 18.2 |
| Grok 4.3 | 17.0 |
| MiniMax M2.7 | 8.4 |