Text-to-Video Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
73.89Semantic Score
6
Feb 26, 2026
60.43VQAA
6
Feb 26, 2026
0.262HPS (V) Score
6
Feb 26, 2026
82Instruction Following
6
Feb 26, 2026
61.2Human Preference Score
6
Feb 26, 2026
93.09Background Consistency
5
Jun 15, 2026
4.84Stability
5
Jun 11, 2026
0FVD
5
May 29, 2026
58.01Physics Reasoning Accuracy
5
May 22, 2026
78.49VBench Quality Score
5
May 22, 2026
43.42Latency (s)
5
May 21, 2026
284.71HD-FVD
5
May 19, 2026
69.53Visual Quality Score
5
Apr 21, 2026
32.69Human Score
5
Apr 8, 2026
48.17Human Score
5
Apr 8, 2026
30.364CLIP-SCORE
5
Apr 7, 2026
31.445CLIP-SCORE
5
Apr 7, 2026
31.997CLIP-SCORE
5
Apr 7, 2026
32.93FID-VID
5
Mar 24, 2026
0.665ImgQual
5
Mar 11, 2026
0.665ImgQual
5
Mar 11, 2026
5.56FLOPs Speedup
5
Feb 26, 2026
91.45Dino Score (10s)
5
Feb 26, 2026
0.46LPIPS
5
Feb 26, 2026
11.2Latency (s)
5
Feb 26, 2026