bytedance / sdxl-lightning-4step

SDXL-Lightning is a lightning-fast text-to-image generation model that produces high-quality 1024px images in just a few steps, distilled from Stable Diffusion XL.

A100

0.0026999

GitHub

License

Weights

Paper

Input

Prompt

Input prompt

Negative Prompt

Negative Input prompt

Width

Min: 256 - Max: 1280

Width of output image. Recommended 1024 or 1280

Height

Min: 256 - Max: 1280

Height of output image. Recommended 1024 or 1280

Additional Settings

Customize your input with more control.

Num Outputs

Min: 1 - Max: 4

Number of images to output.

scheduler

Num Inference Steps

Min: 1 - Max: 10

Number of denoising steps. 4 for best results

Guidance Scale

Min: 0 - Max: 50

Scale for classifier-free guidance

Seed

Random seed. Leave blank to randomize the seed

Disable Safety Checker

Disable safety checker for generated images

You need to be logged in to run this model and view results.

Output

{
  "error": "",
  "inferenceTime": 1938,
  "output": [
    "https://media.modelrunner.ai/5hHDsfs2OKgtehEVy72MD.png"
  ],
  "input": {
    "seed": 103,
    "width": 1280,
    "height": 1024,
    "prompt": "two friends cooking together in a sunlit Scandinavian kitchen, one slicing herbs while the other stirs a pot, visible hands, natural food textures, gentle steam, depth of field, candid lifestyle photograph, 35mm, morning light through window, photoreal",
    "scheduler": "K_EULER",
    "num_outputs": 1,
    "guidance_scale": 0,
    "negative_prompt": "worst quality, low quality, deformed, extra fingers, fused hands, disembodied limbs, awkward pose, motion blur, watermark, logo",
    "num_inference_steps": 4,
    "disable_safety_checker": false
  },
  "logs": "Using seed: 103\nPrompt: two friends cooking together in a sunlit Scandinavian kitchen, one slicing herbs while the other stirs a pot, visible hands, natural food textures, gentle steam, depth of field, candid lifestyle photograph, 35mm, morning light through window, photoreal\n  0%|          | 0/4 [00:00<?, ?it/s]\n 50%|█████     | 2/4 [00:00<00:00, 14.51it/s]\n100%|██████████| 4/4 [00:00<00:00, 14.19it/s]\n100%|██████████| 4/4 [00:00<00:00, 14.23it/s]\n"
}

Generated in 1.938 seconds

Logs (7 lines)

Examples

Model Details

# SDXL-Lightning

**SDXL-Lightning** is a state-of-the-art text-to-image generation model designed for both **speed** and **quality**, capable of producing **high-resolution 1024px images** in remarkably few steps.

Built upon the foundation of [`stabilityai/stable-diffusion-xl-base-1.0`](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0), it leverages *Progressive Adversarial Diffusion Distillation* to achieve rapid yet powerful image synthesis.

## Features - 🚀 **Lightning-fast inference** with 1-step, 2-step, 4-step, and 8-step distilled models. - **2-step, 4-step, and 8-step** models deliver outstanding image quality. - **1-step** model is experimental, exploring the limits of extreme efficiency.

- 🧩 **Model formats**: - **Full UNet checkpoints** – deliver the best image quality. - **LoRA checkpoints** – lightweight and easily applied to other base models.

## Open Source We release SDXL-Lightning fully open-sourced as part of our research to empower the community with cutting-edge fast diffusion technology. For more details, please see our paper: **SDXL-Lightning: Progressive Adversarial Diffusion Distillation**.

bytedance / sdxl-lightning-4step

Model Input

Input

Additional Settings

Model Output

Output

Model Example Requests

Examples

Model Details

Model Details