Instruction Following on MaXIFE
90.58Loose Score (MaXIFE)Gemini 3.1 Pro
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Gemini 3.1 ProThinking mode=true2026.05 | 90.58 | 87.62 | 89.1 | |
| GPT-5.5Thinking mode=true2026.05 | 89.96 | 84.58 | 87.27 | |
| Qwen3.5-397B-A17BThinking mode=true2026.05 | 89.88 | 86.8 | 88.34 | |
| Kimi K2.6Thinking mode=true2026.05 | 89.19 | 85.8 | 87.5 | |
| GLM5.1Thinking mode=true2026.05 | 88.59 | 85.61 | 87.1 | |
| DeepSeek-V4-ProThinking mode=true2026.05 | 88.53 | 84.14 | 86.34 | |
| GPT-5.5Thinking mode=false2026.05 | 87.04 | 81.04 | 84.04 | |
| Kimi K2.6Thinking mode=false2026.05 | 86.4 | 80.49 | 83.45 | |
| GLM5.1Thinking mode=false2026.05 | 86.4 | 80.49 | 83.45 | |
| Qwen3.5-397B-A17BThinking mode=false2026.05 | 86.26 | 80.46 | 83.36 | |
| DeepSeek-V4-ProThinking mode=false2026.05 | 84.98 | 78.77 | 81.88 | |
| Gemma4-26B-A4BThinking mode=false2026.05 | 84.87 | 79.25 | 82.06 | |
| Gemma4-E4BThinking mode=false2026.05 | 81.58 | 74.83 | 78.21 | |
| Qwen3.6-35B-A3BThinking mode=false2026.05 | 80.72 | 73.39 | 77.06 | |
| Hy-MT2-30B-A3BThinking mode=false2026.05 | 80.46 | 74.31 | 77.39 | |
| Gemma4-E2BThinking mode=false2026.05 | 77.32 | 68.79 | 73.06 | |
| Hy-MT2-7BThinking mode=false2026.05 | 76.77 | 68.73 | 72.75 | |
| Hy-MT2-1.8BThinking mode=false2026.05 | 61.29 | 51.05 | 56.17 |