Single-turn child safety evaluation on KIDBench single-turn
4.98Safety ScoreQwen 3.6-27B
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen 3.6-27BCue Setting=Explicit Age2026.05 | 4.98 | 4.98 | 4.99 | 4.98 | 4.97 | 4.98 | |
| DeepSeek V4 FlashCue Setting=Explicit Age2026.05 | 4.92 | 4.92 | 4.9 | 4.88 | 4.73 | 4.87 | |
| Gemini 3.1 Flash LiteCue Setting=Explicit Age2026.05 | 4.87 | 4.9 | 4.93 | 4.86 | 4.78 | 4.87 | |
| Gemma 4-31BCue Setting=Explicit Age2026.05 | 4.86 | 4.9 | 4.9 | 4.82 | 4.76 | 4.85 | |
| GPT 5 MiniCue Setting=Explicit Age2026.05 | 4.83 | 4.88 | 4.9 | 4.9 | 4.83 | 4.87 | |
| Gemma 3-12BCue Setting=Explicit Age2026.05 | 4.82 | 4.86 | 4.93 | 4.82 | 4.71 | 4.83 | |
| Claude Haiku 4.5Cue Setting=Explicit Age2026.05 | 4.76 | 4.84 | 4.85 | 4.71 | 4.63 | 4.76 | |
| Qwen 3.6-27BCue Setting=Implicit Cue2026.05 | 4.68 | 3.81 | 4.41 | 4.46 | 4.1 | 4.29 | |
| Qwen 3-8BCue Setting=Explicit Age2026.05 | 4.67 | 4.75 | 4.84 | 4.64 | 4.48 | 4.68 | |
| Qwen 3.5-4BCue Setting=Explicit Age2026.05 | 4.64 | 4.63 | 4.77 | 4.57 | 4.44 | 4.61 | |
| Llama 3.3-70BCue Setting=Explicit Age2026.05 | 4.61 | 4.69 | 4.76 | 4.54 | 4.35 | 4.59 | |
| Gemini 3.1 Flash LiteCue Setting=Implicit Cue2026.05 | 4.57 | 3.96 | 4.38 | 4.36 | 4.11 | 4.27 | |
| DeepSeek V4 FlashCue Setting=Implicit Cue2026.05 | 4.55 | 4.4 | 4.55 | 4.48 | 4.17 | 4.43 | |
| Gemma 4-31BCue Setting=Implicit Cue2026.05 | 4.49 | 3.94 | 4.17 | 4.14 | 3.97 | 4.14 | |
| Claude Haiku 4.5Cue Setting=Implicit Cue2026.05 | 4.37 | 3.82 | 4.08 | 4.11 | 3.63 | 4 | |
| GPT 5 MiniCue Setting=Implicit Cue2026.05 | 4.35 | 3.57 | 4.1 | 4.13 | 3.79 | 3.99 | |
| Qwen 3.5-4BCue Setting=Implicit Cue2026.05 | 4.31 | 3.54 | 4.03 | 3.96 | 3.66 | 3.9 | |
| Gemma 3-12BCue Setting=Implicit Cue2026.05 | 4.28 | 3.45 | 4.28 | 4.08 | 3.8 | 3.98 | |
| Llama 3.1-8BCue Setting=Explicit Age2026.05 | 4.28 | 4.1 | 4.13 | 3.85 | 3.68 | 4.01 | |
| Gemma 3-4BCue Setting=Explicit Age2026.05 | 4.27 | 4.49 | 4.72 | 4.29 | 4.21 | 4.4 | |
| Llama 3.3-70BCue Setting=Implicit Cue2026.05 | 4.25 | 3.56 | 3.93 | 3.73 | 3.32 | 3.76 | |
| Qwen 3.6-27BCue Setting=No Cue2026.05 | 4.16 | 2.31 | 2.97 | 3.36 | 2.87 | 3.14 | |
| DeepSeek V4 FlashCue Setting=No Cue2026.05 | 4.15 | 2.8 | 3.18 | 3.47 | 2.97 | 3.31 | |
| Llama 3.2-3BCue Setting=Explicit Age2026.05 | 4.02 | 3.97 | 4.2 | 3.67 | 3.51 | 3.87 | |
| Qwen 3-8BCue Setting=Implicit Cue2026.05 | 3.99 | 3.23 | 3.92 | 3.65 | 3.17 | 3.59 | |
| GPT 5 MiniCue Setting=No Cue2026.05 | 3.98 | 2.33 | 2.91 | 3.28 | 2.79 | 3.06 | |
| Qwen 3.5-4BCue Setting=No Cue2026.05 | 3.97 | 2.15 | 2.66 | 3.15 | 2.77 | 2.94 | |
| Llama 3.3-70BCue Setting=No Cue2026.05 | 3.92 | 2.57 | 3.11 | 3.3 | 2.82 | 3.14 | |
| Gemma 4-31BCue Setting=No Cue2026.05 | 3.92 | 2.14 | 2.47 | 2.9 | 2.66 | 2.82 | |
| Qwen 3-8BCue Setting=No Cue2026.05 | 3.9 | 2.23 | 2.75 | 3.18 | 2.6 | 2.93 | |
| Claude Haiku 4.5Cue Setting=No Cue2026.05 | 3.87 | 2.65 | 2.85 | 3.11 | 2.69 | 3.03 | |
| Gemini 3.1 Flash LiteCue Setting=No Cue2026.05 | 3.86 | 2.26 | 2.73 | 3.02 | 2.71 | 2.92 | |
| Llama 3.1-8BCue Setting=Implicit Cue2026.05 | 3.86 | 2.78 | 3.14 | 3.09 | 2.71 | 3.11 | |
| Llama 3.1-8BCue Setting=No Cue2026.05 | 3.85 | 2.24 | 2.66 | 2.86 | 2.45 | 2.81 | |
| Gemma 3-4BCue Setting=Implicit Cue2026.05 | 3.79 | 2.96 | 4.09 | 3.59 | 3.19 | 3.52 | |
| Gemma 3-12BCue Setting=No Cue2026.05 | 3.75 | 1.97 | 2.83 | 3.05 | 2.8 | 2.88 | |
| Llama 3.2-3BCue Setting=No Cue2026.05 | 3.65 | 2.27 | 2.71 | 2.84 | 2.46 | 2.79 | |
| Llama 3.2-3BCue Setting=Implicit Cue2026.05 | 3.57 | 2.73 | 3.3 | 2.95 | 2.6 | 3.03 | |
| Gemma 3-4BCue Setting=No Cue2026.05 | 3.55 | 1.93 | 2.84 | 2.85 | 2.52 | 2.74 |