Multi-turn instruction following on MultiIF
68.93Normalized ScoreQwen3-Max-Thinking
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen3-Max-Thinkingformatting=multi-turn, score_aggregation=stricter, setup=normalized2026.03 | 68.93 | 43 | |
| DeepSeek-V3.2formatting=multi-turn, score_aggregation=stricter, setup=normalized2026.03 | 64.23 | 43 | |
| MiniMax-M2.5formatting=multi-turn, score_aggregation=stricter, setup=normalized2026.03 | 31.76 | 43 | |
| Kimi-K2.5formatting=multi-turn, score_aggregation=stricter, setup=normalized2026.03 | 26.94 | 43 | |
| GLM-5formatting=multi-turn, score_aggregation=stricter, setup=normalized2026.03 | 26.25 | 43 |