-
ControlNet has changed Stable Diffusion forever. It lets you copy human poses, color, and content of a reference image, to name a few. Mastering ControlNet is a logical next step after mastering prompt building.
-
OpenPose Editor allows you to edit the control image from ControlNet‘s Openpose preprocessor. Now you can adjust the pose precisely when using ControlNet.
-
After Detailer automatically fix faces and hands. It greatly increases the chance of getting a good image.
-
Style Selector for SDXL conveniently adds preset keywords to prompts and negative prompts to achieve certain styles. The styles available are the same as those in DreamStudio, the official Stable Diffusion image generator.
-
The best resolution for Stable Diffusion v1 is 512 pixels on one side. Ultimate SD Upscale allows you to enlarge images with ControlNet and tile by tile. It is helpful if you don’t have a GPU card with a large VRAM.
-
Fast and Simple Face Swap Extension for StableDiffusion WebUI (A1111, SD.Next, Cagliostro)
-
Deforum is a one-of-a-kind video that you can make using Stable Diffusion. It looks psychedelic and is very fun to watch.
-
AnimateDiff lets you make short videos using prompts with any Stable Diffusion model.
-
The CLIP interrogator lets you query the prompt of an image. Although you can use the Interrogate CLIP function on the Img2img page to find a prompt of an image, this extension gives you more options. You can specify which language model to use and analyze an image with different databases.
-
You have many images. Want to find the older ones? Instead of using File Explorer (Windows) or Finder (Mac), you can view and search them directly in AUTOMATIC1111 using the Infinite Image Browser. Plus, you can search for images based on prompts and models. The Image Browser is especially useful when accessing A1111 from another machine, where browsing images is not easy.
-
If you use Stable Diffusion, you probably have downloaded a model from Civitai. But you must ensure putting the checkpoint, LoRA, and textual inversion models in the right folders. Civitai Helper lets you download models from Civitai right in the AUTOMATIC1111 GUI. As a bonus, the cover image of the models will be downloaded. Use the Scan button to download the model images of existing models.
-
SDXL-TURBO > sd_xl_turbo_1.0.safetensors [13.9 GB]
-
SDXL-TURBO(fp16) > sd_xl_turbo_1.0_fp16.safetensors [6.94 GB]
-
SDXL BASE 1.0 > sd_xl_base_1.0.safetensors [6.94 GB]
-
SDXL 1.0 REFINER > sd_xl_refiner_1.0.safetensors [6.08 GB]
-
SDXL 1.0 VAE > sdxl_vae.safetensors [335 MB]
-
SD-v1.5 > v1-5-pruned-emaonly.safetensors [4.27 GB]
-
SD-v1.5-inpaint > sd-v1-5-inpainting.ckpt [4.27 GB]
-
SD-v2.1-768 > v2-1_768-ema-pruned.safetensors [5.21 GB]
-
SD-v2.1-512 > v2-1_512-ema-pruned.safetensors [5.21 GB]
-
SD-v1.5 > ControlNet Collection
-
SDXL 1.0 > ControlNet Collection
NOTE: A1111-ControlNet-Path: stable-diffusion-webui\extensions\sd-webui-controlnet\models