whatmodel.sh / image / SDXL-Turbo
SDXL-Turbo
↗ share
+ contribute
SDXL-Turbo is an open-weights image model from Stability AI, released November 2023. Its published size is 2.57B parameters. Licence: sai-nc-community.
Catalogue identity notes
stabilityai/sdxl-turbo (distilled from SDXL 1.0). Memory estimated as for SDXL 1.0.
Model information sources
UNet only: HF Hub safetensors total for the official repo (pinned revision), 2,567,463,684 = unet/diffusion_pytorch_model.safetensors. The two CLIP text encoders and the VAE are not included. Observed 2026-10-07.
Score within the LMArena text-to-image comparison set only.
$ how it compares
The image models ranked either side of SDXL-Turbo, out of 197 with a score, then more from Stability AI.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 168 |  LCM-SDXL | 34±3 | local | ||
| 169 |  OmniGen (v1) | 34±4 | local | ||
| 170 | SDXL-Turbo | 34±2 | local | ||
| 171 |  IF-L | 33±3 | local | ||
| 172 |  Lumina-DiMOO | 32±4 | local | ||
| more from Stability AI | |||||
| 137 | Stable Diffusion 3.5 Large | 42±1 | api $0.065/image +local | ||
| 141 | Stable Diffusion 3.5 Medium | 41±3 | api $0.035/image +local | ||
| 148 | Stable Diffusion XL 0.9 (base) | 40±6 | local | ||
$ what it's good at
text to image▂▅▅▃▄▄█▃▁▂▁▁ #105 of 126 · 89% of its score · 3 tests
Beats Latent Diffusion text2img-large (LDM-KL-8-G, LAION-400M), VQ-Diffusion (ITHQ), IF-XL here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| GenAI-Arena text-to-image (Elo, 4 Jun 2024) | 935.0 ✎ | ▃▃··▃·▃·█·▃▃ beats 2/8 | 33% |
| EvalMuse-40K model ranking (FGA-BLIP2 overall alignment) | 3.15 / 5 ✎ | █▄·▄██▄▄█▄▄▄ beats 9/14 | 31% |
| ConceptMix (k=2 full-mark score) | 35.0% ✎ | ▄·▄·█▄▄▄···▄ beats 2/7 | 25% |
other tests1 not in a task · 10% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| MS-COCO 10K FID (whole image) ↓ (SDXL-Lightning protocol) | 23.7100 FID ✎ | █··██······█ beats 0/3 | 10% |
$ where to run it
Other ways to run it
version
- Run it as published · needs ~8.9 GB, NVIDIA only · from huggingface.co/stabilityai (official)
uv run --with diffusers --with transformers --with accelerate --with torch python -c "import torch; from diffusers import AutoPipelineForText2Image; p=AutoPipelineForText2Image.from_pretrained('stabilityai/sdxl-turbo', torch_dtype=torch.float16, variant='fp16').to('cuda'); p(prompt='a cat holding a sign', num_inference_steps=1, guidance_scale=0.0).images[0].save('out.png')"
Official weights and documentation: huggingface.co/stabilityai ↗
$ how the score was worked out · GenAI Arena: An Open Evaluati… 33%, EvalMuse-40K: A Reliable and… 31%, ConceptMix: A Compositional I… 25% and 1 more
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 GenAI Arena: An Open Evaluation Platform for Generative Models (arXiv v3) | 1 | 33% |
| EvalMuse-40K: A Reliable and Fine-Grained Benchmark with Comprehensive Human Annotations for Text-to-Image Generation Model Evaluation (arXiv v2) | 1 | 31% |
| ConceptMix: A Compositional Image Generation Benchmark with Controllable Difficulty (arXiv v1) | 1 | 25% |
| SDXL-Lightning: Progressive Adversarial Diffusion Distillation (arXiv:2402.13929v3) | 1 | 10% |
| prior · pull toward the middle | 2% | |
| weighted average | = 33.5 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json