Hallucination Evaluation on MVH-Bench Cross-View benchmark leaderboard · SOTA2 Research

Hallucination Evaluation on MVH-Bench Cross-View

71Binary AccuracyGemini 2.5 Flash

Evaluation Results

Method	Links
Gemini 2.5 Flash2026.03		71	51.16	28.06	69.54	48.89	268.65
Gemini 2 Flash2026.03		69.91	44.54	23.24	73.06	53.8	264.55
Claude 4.5 Sonnet2026.03		69.26	49.35	23.61	72.03	49.42	263.67
GPT-4o2026.03		68.87	46.48	23.43	55.74	32.31	226.83
Qwen2.5-VL-7B2026.03		67.45	47.45	21.95	70.7	50.56	258.11
Qwen2-VL-7B2026.03		61.97	37.92	14.26	58.71	31.3	204.16
GPT-4o mini2026.03		61.92	35.83	14.72	61.06	33.06	206.59
LLaVA-NeXT-I-7B2026.03		61.02	37.31	15.09	62.69	38.33	214.44
Mantis-8B-Idefics22026.03		58.94	35.05	11.76	58.93	33.7	198.38
Gemma3-12B2026.03		58.35	24.91	5.09	58.89	30.37	177.61
InternVL3-14B2026.03		55.37	29.03	4.48	56.9	26.11	171.89
LLaVA-OneVision-7B2026.03		52.36	27.17	4.26	59.26	29.72	172.77
InternVL3-8B2026.03		50.86	22.69	1.76	50.93	15.74	141.98
Deepseek-VL-Chat-7B2026.03		50.28	20.14	0.93	49.31	8.24	128.89
Llama-3.2-11B-V-I2026.03		49.38	18.01	1.94	47.41	16.11	132.85