Qwen-Image on DGX Spark: VRAM, speed and fit
Qwen-Image at 16-bit runs on the DGX Spark 128 GB, about 9.2 min per image.
Image generation · 1328 × 1328
Runs comfortably · on DGX Spark 128 GB · 16-bit (BF16)
- GPU memory
- 57.3 GB
- of 128 GB
- Time
- 9.2 min
- per image
Where the memory goes
- Model
- 45.8 GB
- Working
- 3.0 GB
- Overhead
- 819 MB
- OS
- 7.7 GB
- Free
- 70.7 GB
- Memory
- 57.3 GB of 128 GB
- Output
- 1328 × 1328 · 50 steps
- Time
- 9.2 min per image
- Running cost
- $4.73/mo at home · $72/mo on a cloud A40
Validation: Not validated
| Check | Target | Result | Status |
|---|---|---|---|
| Fits in GPU memory | 120 GB usable | 49.6 GB needed | Pass |
| Headroom | 5% free | 70.7 GB free (59%) | Pass |
| Time per image | under 30 s | 9.2 min | Fails |
Targets: an image in under 30 s, and 5% of memory left free. Calculated from published specs, not benchmarked.
The setup
- Model
- Qwen-Image
- Parameters
- 20.4B
- Precision
- 16-bit (BF16)
- Output
- 1328 × 1328
- Steps
- 50
- Hardware
- DGX Spark 128 GB
- GPU can use
- 120 GB of 128 GB
Running cost
- At home
- $4.73/mo
- 216 W at the wall, $0.006 an image
- Renting instead
- $72/mo
- A40 at $0.59/h, $0.04 an image
4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 21, 2026.
Where to get it
Qwen-Image on similar machines
- M5 Ultra 512 GBRuns
- Ryzen AI Max+ 395 64 GBRuns
- Ryzen AI Max+ 395 128 GBRuns
- RTX SparkRuns
- DGX Station GB300Runs
All machines: Qwen-Image VRAM requirements. Everything that runs here: what can the DGX Spark run?
Other models that run on DGX Spark
Questions
Can Qwen-Image run on DGX Spark?
Qwen-Image at 16-bit runs on the DGX Spark 128 GB, about 9.2 min per image.
Report VR-1PJP7ZM, calculated Nov 22, 2026 from published specs, not benchmarked.