SVD-XT
stabilityai/stable-video-diffusion-img2vid-xt
Stable Video Diffusion, image-to-video 14 frames.
Parameters
1.5B
Context
8K
Downloads
340K
Likes
1.6K
Architecture
svd
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 11.0 GB RAM 32.0 · Disk 43.4 | 15.0 GB RAM 32.0 · Disk 69.4 | 22.8 GB RAM 34.3 · Disk 108.4 |
| FP88bit | 8.3 GB RAM 32.0 · Disk 41.2 | 12.1 GB RAM 32.0 · Disk 67.2 | 19.8 GB RAM 32.0 · Disk 106.2 |
| INT88bit | 8.3 GB RAM 32.0 · Disk 41.2 | 12.1 GB RAM 32.0 · Disk 67.2 | 19.8 GB RAM 32.0 · Disk 106.2 |
| AWQ4bit | 6.9 GB RAM 32.0 · Disk 40.1 | 10.7 GB RAM 32.0 · Disk 66.1 | 18.4 GB RAM 32.0 · Disk 105.1 |
| GPTQ4bit | 7.0 GB RAM 32.0 · Disk 40.1 | 10.7 GB RAM 32.0 · Disk 66.1 | 18.4 GB RAM 32.0 · Disk 105.1 |
| GGUF4bit | 7.0 GB RAM 32.0 · Disk 40.2 | 10.8 GB RAM 32.0 · Disk 66.2 | 18.4 GB RAM 32.0 · Disk 105.2 |
GPU Recommendations
1× RTX 5080
16 GB VRAM
≈ 1440 video-sec/GPU-h
Effective production throughput ≈ 864 video-sec/GPU-h
score 0.6723
1× Tesla V100
16 GB VRAM
≈ 1440 video-sec/GPU-h
Effective production throughput ≈ 864 video-sec/GPU-h
score 0.6676
8× RTX 5090
256 GB VRAM
≈ 100 video-sec/GPU-h
Effective production throughput ≈ 60 video-sec/GPU-h
score 0.3376
1× RTX 3090
24 GB VRAM
≈ 42.4 video-sec/GPU-h
Effective production throughput ≈ 25.4 video-sec/GPU-h
score 0.6142
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
Tesla V100
16 GB VRAM
$35.7
/mo · vast
RTX 3090
24 GB VRAM
$63.8
/mo · vast
RTX 4070S Ti
16 GB VRAM
$68.5
/mo · vast
RTX 5070
12 GB VRAM
$73.1
/mo · vast
L4
24 GB VRAM
$76.2
/mo · vast