diff --git a/Features.md b/Features.md
index 62d82cd..031865b 100644
--- a/Features.md
+++ b/Features.md
@@ -609,7 +609,7 @@ This implementation of optimization does not require any modification to the ori
 With this lightweight VAE enabled via settings, it typically allows for very large, fast generations with a small quality loss.
 This gain can be very large, maximum generations with --lowvram can increase from `1152x1152` to `2560x2560`
 
-![image](https://github.com/AUTOMATIC1111/stable-diffusion-webui/assets/98228077/016fa4be-155d-433b-88b9-00cf7b452a8a)
+![image](images/taesd-vae.png)
 
 
 # Face restoration
@@ -637,7 +637,7 @@ Gradio's loading graphic has a very negative effect on the processing speed of t
 My RTX 3090 makes images about 10% faster when the tab with gradio is not active. By default, the UI now hides loading progress animation and replaces it with static "Loading..." text, which achieves the same effect. Use the `--no-progressbar-hiding` commandline option to revert this and show loading animations.
 
 # Caching Models
-![image](https://github.com/AUTOMATIC1111/stable-diffusion-webui/assets/98228077/aaee8ca8-eb9e-4f22-8b7b-dfce51211e63)
+![image](images/caching-modles.png)
 
 If you want faster swapping between models, increase the counter in settings. Webui will keep models you've swapped from in ram. 
 
diff --git a/How-to-make-your-own-Inpainting-model.md b/How-to-make-your-own-Inpainting-model.md
index 8ed7ff1..f27df4a 100644
--- a/How-to-make-your-own-Inpainting-model.md
+++ b/How-to-make-your-own-Inpainting-model.md
@@ -1,7 +1,7 @@
 Making your own inpainting model is very simple:
 
 
-![screenshot](https://github.com/AUTOMATIC1111/stable-diffusion-webui/assets/40751091/4bdbab38-9237-48ea-9698-a036a5c96585)
+![screenshot](images/making-your-own-inpainting-model.png)
 
 1. Go to Checkpoint Merger
 2. Select "Add Difference"
diff --git a/Optimizations.md b/Optimizations.md
index 39fcb86..f6641a1 100644
--- a/Optimizations.md
+++ b/Optimizations.md
@@ -18,7 +18,7 @@ A number of optimization can be enabled by [commandline arguments](Command-Line-
 | `--upcast-sampling`           | For Nvidia and AMD cards normally forced to run with `--no-half`, [should improve generation speed](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/8782).    
                                                                                                                                                                                                                                                                                                                                       
 As of [version 1.3.0](https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/tag/v1.3.0), `Cross attention optimization` can be selected under settings. xFormers still needs to enabled via `COMMANDLINE_ARGS`.
-![2023-06-21 22_53_54_877 chrome](https://github.com/AUTOMATIC1111/stable-diffusion-webui/assets/40751091/c72576e1-0f51-4643-ad91-e9aaec4fc125)
+![change-cross-attention-optimization](images/change-cross-attention-optimization.png)
 
 
 Extra tips (Windows): 
diff --git a/images/caching-modles.png b/images/caching-modles.png
new file mode 100644
index 0000000..12554d9
Binary files /dev/null and b/images/caching-modles.png differ
diff --git a/images/change-cross-attention-optimization.png b/images/change-cross-attention-optimization.png
new file mode 100644
index 0000000..adfc752
Binary files /dev/null and b/images/change-cross-attention-optimization.png differ
diff --git a/images/making-your-own-inpainting-model.png b/images/making-your-own-inpainting-model.png
new file mode 100644
index 0000000..74a233c
Binary files /dev/null and b/images/making-your-own-inpainting-model.png differ
diff --git a/images/taesd-vae.png b/images/taesd-vae.png
new file mode 100644
index 0000000..d09e104
Binary files /dev/null and b/images/taesd-vae.png differ