Krea 2 Turbo needs about 9.3 GB of memory at Q4_K_M (a 7.9 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 5 quants, from 6.3 GB (Q3_K_M) to 14.3 GB (Q8_0).
Krea 2 Turbo – a 12.8 billion‑parameter diffusion transformer in GGUF format
Krea 2 Turbo is a 12.8 billion‑parameter diffusion transformer quantized with ByteShape’s ShapeLearn. Five GGUF variants range from 6.26 GB to 14.31 GB. All variants generate high‑quality images and run in ComfyUI via the ComfyUI‑GGUF extension.
Good at
- High‑quality image generation
- Fast inference on RTX 5090
- Wide range of quantization options
Limits
- No vision support beyond image generation
- No classifier‑free guidance
- Requires 8 k context for memory estimates
Best for
- Stable diffusion style image creation
- Rapid prototyping of diffusion models
- Testing quantization trade‑offs
Who it’s for. Developers and researchers who need a compact diffusion model for local inference. The model is suitable for users with GPUs that can accommodate the chosen quantization size.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_M | 6.3 GB | 7.5 GB | Tight | Get file → | |
| Q4_K_S | 6.8 GB | 8.2 GB | Too big | Get file → | |
| Q4_K_M | 7.9 GB | 9.3 GB | Too big | Get file → | |
| Q6_K | 11.7 GB | 13.6 GB | Too big | Get file → | |
| Q8_0 | 14.3 GB | 16.5 GB | Too big | Get file → |
For 8 GB VRAM, use Q3_K_M (6.26 GB, ~7.5 GB memory at 8 k context). For 12 GB VRAM, Q4_K_M (7.9 GB, ~9.3 GB memory). For 16 GB VRAM, Q6_K (11.69 GB, ~13.6 GB memory). The Q8_0 variant requires the most memory and is best for 20 GB+ VRAM.
What GPU do you need for Krea 2 Turbo?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: the best quant that fits is Q3_K_M (6.3 GB file, about 7.5 GB in use).
- 12 GB VRAM: the best quant that fits is Q4_K_M (7.9 GB file, about 9.3 GB in use).
- 16 GB VRAM: the best quant that fits is Q6_K (11.7 GB file, about 13.6 GB in use).
- 24 GB VRAM: the best quant that fits is Q8_0 (14.3 GB file, about 16.5 GB in use).
Which Krea 2 Turbo quant should you choose?
Q4_K_M is the usual starting point for Krea 2 Turbo: it balances file size and quality well. Here it is a 7.9 GB file.
The Q8_0 file is 14.3 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q3_K_M at 6.3 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Krea 2 Turbo
How much VRAM does Krea 2 Turbo need?
About 9.3 GB at Q4_K_M with an 8K context window. That is the 7.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Krea 2 Turbo on an 8 GB GPU?
Yes, with Q3_K_M: the 6.3 GB file needs about 7.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Krea 2 Turbo GGUF?
It depends on the quant: Q3_K_M is 6.3 GB, Q4_K_S is 6.8 GB, Q4_K_M is 7.9 GB, Q6_K is 11.7 GB, Q8_0 is 14.3 GB.
Where can I download Krea 2 Turbo GGUF?
The files are on Hugging Face in the byteshape/Krea-2-Turbo-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Krea 2 Turbo use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Krea 2 Turbo on 8 GB VRAM?
Yes, the Q3_K_M variant is 6.26 GB and uses about 7.5 GB at 8 k context, fitting within 8 GB VRAM.
Does Krea 2 Turbo support classifier‑free guidance?
No, the model is step‑distilled with 8 steps and no classifier‑free guidance.
What backend is required to run this model?
It runs in ComfyUI via the ComfyUI‑GGUF extension; vLLM‑Omni builds are also available for faster inference.