Stable Diffusion Web UI Cloud Inference

What capabilities does this extension offer?

This extension enables faster image generation without the need for expensive GPUs and seamlessly integrates with the AUTOMAIC1111 UI.

Benefits:

1. No expensive GPUs required, can even use the CPU.
2. No need to change your workflow, compatible with the usage and scripts of sd-webui, such as X/Y/Z Plot, Prompt from file, etc.
3. Support for 10000+ Checkpoint models, don't need download

Compatibility and Limitations

Feature	Support	Limitations
txt2img	✅
txt2img_hires.fix	✅
txt2img_sdxl_refiner	✅
txt2img_controlnet	✅
img2img	✅
img2img_inpaint	✅
img2img_sdxl_refiner	✅
img2img_controlnet	✅
extras upscale	✅
vae model	✅
scripts - X/Y/Z plot	✅
scripts - Prompt matrix	✅
scripts - Prompt from file	✅

How it works

Guide

1. Install sd-webui-cloud-inference

2. Get your omniinfer.io Key

Open omniinfer.io in browser

We can choice "Google Login" or "Github Login"

3. Enable Cloud Inference feature

Let us back to Cloud Inference tab of stable-diffusion-webui

4. Test Txt2Img

Let us back to Txt2Img tab of stable-diffusion-webui

From now on, you can give it a try and enjoy your creative journey.

Furthermore, you are welcome to freely discuss your user experience, share suggestions, and provide feedback on our Discord channel.

5. Advanced - Lora

7. Advanced - Img2img Inpainting

8. Advanced - VAE

or you can use the VAE feature with X/Y/Z

9. Advanced - ControlNet

9. Advanced - ControlNet

10. Advanced - Upscale and Hires.Fix

11. Advanced - Model Browser

12. Advanced - Tiny Model

The AUTOMATIC1111 webui loads the model on startup. However, on low-memory computers like the MacBook Air, the performance is suboptimal. To address this, we have developed a stripped-down minimal-size model. You can utilize the following commands to enable it.

its will reduce memory from 4.8G -> 739MB

1. Download tiny model and config to model config.

wget -O ./models/Stable-diffusion/tiny.yaml https://github.com/omniinfer/sd-webui-cloud-inference/releases/download/tiny-model/tiny.yaml
wget -O ./models/Stable-diffusion/tiny.safetensors https://github.com/omniinfer/sd-webui-cloud-inference/releases/download/tiny-model/tiny.safetensors

2. start webui with tiny model --ckpt=/stable-diffusion-webui/models/Stable-diffusion/tiny.safetensors

13. Advanced - SDXL Refiner