SD Turbo needs about 3.4 GB of memory at F16 (a 2.6 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 2.0 GB (F16) to 3.4 GB (Q8_0).
SD Turbo – fast, single‑step text‑to‑image generation
SD Turbo is a distilled version of Stable Diffusion 2.1, trained for real‑time synthesis. It can generate photorealistic images from a text prompt in a single network evaluation. The model is released as a research artifact and supports both non‑commercial and commercial use.
Good at
- Fast single‑step generation
- High image fidelity at low steps
- Research‑grade diffusion model
Limits
- No vision support
- Requires 8k context for memory estimates
- License requires reference to stability.ai
Best for
- Rapid prototyping of text‑to‑image pipelines
- Real‑time image generation demos
- Research on diffusion model distillation
Who it’s for. Developers and researchers looking to experiment with fast diffusion models. The model is suitable for both academic study and commercial deployment, provided the stability.ai license terms are followed.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| F16 | 2.0 GB | 2.8 GB | Fits | Get file → | |
| F16 | 2.6 GB | 3.4 GB | Fits | Get file → | |
| Q8_0 | 3.4 GB | 4.3 GB | Fits | Get file → |
For a 8 GB VRAM budget, use the Q8_0 quantized file (3.37 GB, 4.3 GB memory at 8k context). If more VRAM is available, the F16 quantized files (2.02 GB or 2.61 GB) provide lower memory usage (2.8 GB or 3.4 GB).
What GPU do you need for SD 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: the best quant that fits is F16 (2.6 GB file, about 3.4 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (3.4 GB file, about 4.3 GB in use).
Which SD Turbo quant should you choose?
F16 (2.6 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
If memory is tight, the smallest file is F16 at 2.0 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about SD Turbo
How much VRAM does SD Turbo need?
About 3.4 GB at F16 with an 8K context window. That is the 2.6 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 SD Turbo on an 8 GB GPU?
Yes, with Q8_0: the 3.4 GB file needs about 4.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of SD Turbo GGUF?
It depends on the quant: F16 is 2.0 GB, F16 is 2.6 GB, Q8_0 is 3.4 GB.
Where can I download SD Turbo GGUF?
The files are on Hugging Face in the Green-Sky/SD-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 SD 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 SD Turbo on 8 GB VRAM?
Yes, the Q8_0 quantized file is 3.37 GB and requires about 4.3 GB memory at 8k context, fitting within 8 GB VRAM.
Does SD Turbo support image resolution higher than 512x512?
The model is primarily tested for 512x512 output; higher resolutions may produce less reliable results.
What license applies to SD Turbo?
The model is released under the stability.ai license; refer to https://stability.ai/license for commercial use.