List your company

Joint Audio-Video Generation on JavisBench

26.2Audio-Video Consistency (AV-IB)LTX-2+OmniNFT
9.76814.03418.322.566May 12, 2026

Evaluation results

MethodLinks
LTX-2+OmniNFTSize=19B, Pipeline Approach=T2AV2026.0526.23.3265.7150.2610.1893144.525.7220.269
LTX-2+ GDPOSize=19B, Pipeline Approach=T2AV2026.0523.33.2095.5230.2650.18430.842.822.318.50.412
LTX-2Size=19B, Pipeline Approach=T2AV2026.0523.22.0385.1970.2720.1731.141.222.319.20.569
MMAudioSize=0.1B, Pipeline Approach=T2V+V2A2026.0519.8———0.16—40.718.2150.849
JavisDiT++Size=2.1B, Pipeline Approach=T2AV2026.0519.81.4625.0490.2820.16431.642.418.415.90.832
JavisDiTSize=3.1B, Pipeline Approach=T2AV2026.0519.71.2914.4780.2630.14330.239.117.915.41.039
FoleyCrafterSize=1.2B, Pipeline Approach=T2V+V2A2026.0519.3———0.149—38.318.615.10.952
See&HearSize=0.4B, Pipeline Approach=T2V+V2A2026.0516———0.129—26.314.311.21.099
TempoTknSize=1.3B, Pipeline Approach=T2A+A2V2026.0513.9——0.084—20.5—12.210.31.532
TPoSSize=1.0B, Pipeline Approach=T2A+A2V2026.0512.4——0.201—22.9—12.99.51.493
ReWaSSize=0.6B, Pipeline Approach=T2V+V2A2026.0511———0.123—2810.47.91.071
UniVerse-1Size=6.4B, Pipeline Approach=T2AV2026.0510.41.3574.8390.2720.11130.924.59.87.70.929