Text-to-Video Generation on VBench (Core Consistency and Quality Metrics) benchmark leaderboard · SOTA2 Research

Text-to-Video Generation on VBench (Core Consistency and Quality Metrics)

58.56Aesthetic QualityFP16

Evaluation Results

Method	Links
FP162026.03		58.56	66.1	97.22	48.61	93.26	96.18	24.48	25.25
FP162026.03		58.05	64.52	98.84	54.17	94.41	97.73	41.74	27.34
Semantic Dialect2026.03		57.78	65.18	98.51	55.56	91.81	96.13	44.57	27.78
Semantic Dialect2026.03		57.39	65.36	98.62	51.39	93.04	96.81	45.09	27.83
Block Dialect2026.03		57.26	62.02	98.76	58.33	92.62	97.45	38.39	27.63
NVFP42026.03		57.09	65.39	98.63	54.17	91.96	96.35	40.1	27.85
NVFP42026.03		57.07	64.4	98.55	50	91.04	96.05	44.05	27.81
Block Dialect2026.03		56.83	60.93	98.79	58.33	91.81	97.13	42.19	27.58
ViDiT-Q2026.03		55.5	59.19	98.72	47.22	90.18	95.12	34.75	27.19
Semantic Dialect2026.03		54.68	63.93	97.75	37.5	93.3	95.72	22.1	23.17
MXFP42026.03		52.99	42.47	99.03	45.83	87.61	94.73	37.13	25.88
MXFP42026.03		52.9	42.31	99.01	44.44	87.74	94.69	34.6	25.9
Semantic Dialect2026.03		51.54	61.03	97.73	51.39	90.88	95.02	18.08	22.55
NVFP42026.03		51	55.21	97.63	47.22	89.18	95.44	19.79	23.15
Block Dialect2026.03		48.26	50.27	98.46	65.28	89.04	94.77	13.84	21.54
NVFP42026.03		46.68	48.3	98.32	40.28	90.24	95.75	11.16	20.04
Block Dialect2026.03		42.43	47.99	98.65	77.78	87.63	94.78	8.26	18.58
MXFP42026.03		34.85	38.27	98.86	88.89	87.72	94.46	5.51	13.3
MXFP42026.03		32.76	30.08	99.04	81.94	86.79	94.33	4.84	13.22
Q-VDiT2026.03		29.37	54.97	85.27	80.56	97.02	97.38	0	0.96
ViDiT-Q2026.03		28.66	64.34	82.64	38.89	91.77	95.25	0	6.5