Instruction Following on IFEval (score)
96.3IFEval ScoreGemini 3.1 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 3.1 ProThinking mode=true2026.05 | 96.3 | |
| Kimi K2.6Thinking mode=true2026.05 | 95.56 | |
| GLM5.1Thinking mode=true2026.05 | 94.09 | |
| GPT-5.5Thinking mode=true2026.05 | 93.9 | |
| GPT-5.5Thinking mode=false2026.05 | 91.68 | |
| DeepSeek-V4-ProThinking mode=true2026.05 | 91.31 | |
| GLM5.1Thinking mode=false2026.05 | 90.57 | |
| Kimi K2.6Thinking mode=false2026.05 | 90.02 | |
| DeepSeek-V4-ProThinking mode=false2026.05 | 89.83 | |
| Qwen3.5-397B-A17BThinking mode=true2026.05 | 89.83 | |
| Gemma4-26B-A4BThinking mode=false2026.05 | 89.8 | |
| Hy-MT2-30B-A3BThinking mode=false2026.05 | 89.8 | |
| Qwen3.5-397B-A17BThinking mode=false2026.05 | 88.54 | |
| Hy-MT2-7BThinking mode=false2026.05 | 86.14 | |
| Gemma4-E4BThinking mode=false2026.05 | 85.76 | |
| Qwen3.6-35B-A3BThinking mode=false2026.05 | 83 | |
| Gemma4-E2BThinking mode=false2026.05 | 80.44 | |
| Hy-MT2-1.8BThinking mode=false2026.05 | 80.22 |