Planning on PlanningBench 1.0 (test)
63.17All-pass RateGPT-5.4-xhigh
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-5.4-xhighType=Closed-source2026.05 | 63.17 | 92.35 | |
| GPT-5.4-highType=Closed-source2026.05 | 58.56 | 84.6 | |
| GPT-5.4-mediumType=Closed-source2026.05 | 58.09 | 90.03 | |
| Gemini-3-1-proType=Closed-source2026.05 | 53.25 | 88.36 | |
| Seed-2.0-pro-highType=Open-source2026.05 | 44.33 | 84.02 | |
| DeepSeek-V3.2-thinkingType=Open-source2026.05 | 37.13 | 80.21 | |
| Hy3-PreviewType=Open-source2026.05 | 36.17 | 78.14 | |
| Qwen3.5-plus-thinkingType=Open-source2026.05 | 34.03 | 77.1 | |
| DeepSeek-R1Type=Open-source2026.05 | 33.07 | 74.04 | |
| Gemini-2.5-proType=Closed-source2026.05 | 31.7 | 76.41 | |
| Seed-2.0-pro-mediumType=Open-source2026.05 | 30.63 | 77.09 | |
| DeepSeek-V3.2-expType=Open-source2026.05 | 29.27 | 76.69 | |
| Hy2.1Type=Open-source2026.05 | 24.61 | 71.18 | |
| Seed-1.8Type=Open-source2026.05 | 22.4 | 71.69 | |
| Qwen3-30b-moeType=Open-source2026.05 | 12.15 | 57.4 | |
| Qwen3-32bType=Open-source2026.05 | 0.27 | 30.11 | |
| Qwen3-14bType=Open-source2026.05 | 0 | 25.69 | |
| Qwen3-8bType=Open-source2026.05 | 0 | 22.79 |