List your company

Compositional Hallucination Evaluation on OmniVCHall Real

95AccuracyHuman
41.9655.7369.583.27Jan 31, 2026

Evaluation results

MethodLinks
HumanSize=-2026.0195
Gemini-3-proSize=-2026.0176
GPT-5.2Size=-2026.0174
Qwen3-VL-ThinkingSize=235B2026.0173
Doubao-Seed-1.8Size=-2026.0173
GPT-5Size=-2026.0172
Qwen3-VL-ThinkingSize=32B2026.0171
Qwen3-VL-InstructSize=235B2026.0171
Gemini-2.5-flashSize=-2026.0171
Gemini-2.5-proSize=-2026.0171
Doubao-Seed-1.6Size=-2026.0171
GPT-4oSize=-2026.0170
Qwen3-VL-InstructSize=32B2026.0169
Kimi-VL-ThinkingSize=16B2026.0168
Qwen3-VL-ThinkingSize=4B2026.0167
Qwen3-VL-ThinkingSize=8B2026.0167
Qwen3-VL-InstructSize=8B2026.0165
Molmo2Size=8B2026.0165
Qwen3-VL-InstructSize=4B2026.0164
MiniCPM-V-4.5Size=8B2026.0164
Kimi-VL-InstructSize=16B2026.0164
Molmo2Size=4B2026.0163
Qwen3-VL-ThinkingSize=2B2026.0162
InternVL3.5Size=30B2026.0162
Qwen2.5-VL-InstructSize=32B2026.0162
InternVL3.5Size=4B2026.0161
Qwen2.5-VL-InstructSize=7B2026.0160
InternVL3.5Size=8B2026.0160
Qwen3-VL-InstructSize=2B2026.0159
VideoLLaMA3Size=7B2026.0159
VideoChat-FlashSize=7B2026.0158
MiniCPM-V-4Size=4B2026.0157
VideoChat-FlashSize=2B2026.0155
Qwen2.5-VL-InstructSize=3B2026.0152
VideoLLaMA3Size=2B2026.0151
GLM-4.6vSize=108B2026.0149
LLaVA-NeXT-VideoSize=34B2026.0147
GLM-4.5vSize=108B2026.0147
LLaVA-NeXT-VideoSize=7B2026.0144
GLM-4.6v-flashSize=108B2026.0144