diff --git a/scripts/find_duplicate_images.py b/scripts/find_duplicate_images.py
new file mode 100644
index 0000000..ef620f3
--- /dev/null
+++ b/scripts/find_duplicate_images.py
@@ -0,0 +1,64 @@
+#!/usr/bin/env python3
+import os
+import re
+import argparse
+
+def get_all_images(images_dir):
+    allowed_ext = {'.png', '.jpeg', '.jpg', '.webp', '.gif'}
+    image_files = set()
+    
+    for root, _, files in os.walk(images_dir):
+        for filename in files:
+            ext = os.path.splitext(filename)[1].lower()  # use lower-case
+            if ext in allowed_ext:
+                pth = os.path.abspath(os.path.join(root, filename))
+                image_files.add(pth)
+    return image_files
+
+def get_markdown_image_references(posts_dir):
+    image_refs = set()
+    # regex matches: ![optional alt text](path)
+    pattern = re.compile(r'!\[.*?\]\((.*?)\)')
+    
+    for root, _, files in os.walk(posts_dir):
+        for filename in files:
+            if filename.endswith('.md'):
+                file_path = os.path.join(root, filename)
+                try:
+                    with open(file_path, 'r', encoding='utf-8') as f:
+                        content = f.read()
+                except:
+                    continue
+                matches = pattern.findall(content)
+                for match in matches:
+                    ref = match.strip().strip('"').strip("'")
+                    refreal = os.path.join(root, ref)
+                    #print(refreal, ref, os.path.abspath(refreal))
+                    image_refs.add(os.path.abspath(refreal))
+    return image_refs
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Find unused images (png, jpeg, jpg) that are not referenced by any Markdown post."
+    )
+    parser.add_argument("--docs", required=True,
+                        help="main docs/ directory.")
+    args = parser.parse_args()
+
+    all_images = get_all_images(args.docs)
+
+    image_references = get_markdown_image_references(args.docs)
+    unused_images = all_images - image_references
+
+    if unused_images:
+        print("Unused images:")
+        size_cum = 0
+        for img in sorted(unused_images):
+            size_cum += os.path.getsize(img)
+            print(os.path.relpath(img, start=os.path.abspath(args.docs)))
+        print(f'\nPossible savings: {round(size_cum/1024)} kB')
+    else:
+        print("No unused images found.")
+
+if __name__ == "__main__":
+    main()