Bernini Full — MLLM + Wan2.2 R2V
Reference images (1-5)
Drop Media Here
- or -
Click to Upload
Prompt
Keeping the exact identity and appearance the same as in image0, the person in image0 dances confidently.
Negative prompt
▼
Aspect ratio
16:9
9:16
1:1
4:3
3:4
use first reference image aspect ratio
Seed (0=random)
ZeroGPU budget (0=auto)
↺
0
600
Steps
↺
2
20
Duration (s)
↺
1
15
FPS
RIFE
optimized split (4 HIGH / 2 LOW at 6 steps)
Loras
▼
Guidance (omegas)
▼
Sampler
IPNC
IPNC strength (%)
↺
0
100
Generate
Generated video
Status