Content safety guardrailing on LlavaGuard
87.5Score (%)Yuvion VL
Evaluation Results
| Method | Links | |
|---|---|---|
| Yuvion VLModel Scale=32B, Decoding Mode=Instruct2026.06 | 87.5 | |
| Yuvion VLModel Scale=8B, Decoding Mode=Instruct2026.06 | 81.3 | |
| Yuvion VLModel Scale=8B, Decoding Mode=Reasoning2026.06 | 81 | |
| Yuvion VLModel Scale=32B, Decoding Mode=Reasoning2026.06 | 79.9 | |
| Opus-4.6Model Scale=≈800B2026.06 | 77.6 | |
| Qwen3.5-PlusModel Scale=397B2026.06 | 76.3 | |
| GPT-5.4Model Scale=≈1.5T2026.06 | 75.2 | |
| Qwen3-VLModel Scale=32B, Decoding Mode=Instruct2026.06 | 74.8 | |
| K2.5Model Scale=≈1T2026.06 | 74.1 | |
| Qwen3-VLModel Scale=32B, Decoding Mode=Thinking2026.06 | 71.9 | |
| Qwen3-VLModel Scale=8B, Decoding Mode=Thinking2026.06 | 68.5 | |
| Qwen3-VLModel Scale=8B, Decoding Mode=Instruct2026.06 | 66 |