How much VRAM does Krea 2 Turbo need?
Krea 2 Turbo needs about 13.6 GB at 8-bit, or 8.4 GB as a 4-bit GGUF, at its default output size.
- Type
- Image model
- Parameters
- 12.8B
- Text encoder
- Qwen3-VL-4B
- Released
- Jun 2026
- License
- Krea 2 Community License
Memory at each precision
| Precision | Weights | Peak at 1024 × 1024 |
|---|---|---|
| 16-bit BF16 | 23.9 GB | 25.1 GB |
| 8-bit FP8 · Q8_0 | 12.4 GB | 13.6 GB |
| 5-bit GGUF Q5_K_M | 8.5 GB | 9.7 GB |
| 4-bit GGUF Q4_K_M · NVFP4 | 7.2 GB | 8.4 GB |
The Qwen3-VL-4B text encoder (4.1 GB at FP8) runs first and then waits in system RAM.
What runs it
| Machine | Best fit | Per image |
|---|---|---|
| RTX 3060 12 GB | 5-bit | 81 s |
| RTX 4060 Ti 16 GB | 8-bit | 41 s |
| RTX 3090 | 8-bit | 26 s |
| RTX 4090 | 8-bit | 12 s |
| RTX 5090 | 16-bit | 9.4 s |
| RTX PRO 6000 Blackwell | 16-bit | 4.5 s |
| M4 Max 128 GB | 16-bit | 1.9 min |
| M5 Max 128 GB | 16-bit | 34 s |
| Ryzen AI Max+ 395 128 GB | 16-bit | 52 s |
| DGX Spark 128 GB | 16-bit | 29 s |
| M3 Ultra 256 GB | 16-bit | 66 s |
| M5 Ultra 512 GB | 16-bit | 21 s |
| H100 SXM | 16-bit | 2.8 s |
| DGX Station GB300 | 16-bit | 1.7 s |
1024 × 1024, 8 steps. Times are rough, plus or minus 30%.
Where to get it
Questions
How much VRAM does Krea 2 Turbo need?
Krea 2 Turbo needs about 13.6 GB at 8-bit, or 8.4 GB as a 4-bit GGUF, at its default output size.
Can Krea 2 Turbo run on an RTX 4090?
Yes. At 8-bit it takes about 12 s per image.