diff --git a/haruhi_dl/extractor/embetty.py b/haruhi_dl/extractor/embetty.py
new file mode 100644
index 000000000..4e3882006
--- /dev/null
+++ b/haruhi_dl/extractor/embetty.py
@@ -0,0 +1,54 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    extract_attributes,
+)
+
+
+class EmbettyIE(InfoExtractor):
+    @staticmethod
+    def _extract_entries(webpage, **kw):
+        results = []
+
+        twitter_matches = re.finditer(r'<embetty-tweet\s+[^>]*\bstatus="(\d{18})"[^>]*>', webpage)
+        for match in twitter_matches:
+            results.append({
+                '_type': 'url',
+                'url': 'https://twitter.com/i/web/status/%s' % match.group(1),
+                'ie_key': 'Twitter',
+            })
+
+        video_matches = re.findall(r'<embetty-video\s+[^>]+>', webpage)
+        for match in video_matches:
+            attr = extract_attributes(match)
+            if any(key not in attr for key in ('type', 'video-id')):
+                continue    # invalid
+            service = attr['type'].lower()
+            if service == 'youtube':
+                results.append({
+                    '_type': 'url',
+                    'url': 'https://www.youtube.com/watch?v=%s&t=%s' % (attr['video-id'], attr.get('start-at', '0')),
+                    'ie_key': 'Youtube',
+                })
+            elif service == 'vimeo':
+                results.append({
+                    '_type': 'url',
+                    'url': 'https://vimeo.com/%s#t=%s' % (attr['video-id'], attr.get('start-at', '0')),
+                    'ie_key': 'Vimeo',
+                })
+            elif service == 'facebook':
+                results.append({
+                    '_type': 'url',
+                    'url': 'https://www.facebook.com/video.php?v=%s' % (attr['video-id']),
+                    'ie_key': 'Facebook',
+                })
+
+        return results
+
+        @staticmethod
+        def _extract_urls(webpage, **kw):
+            return [embed['url'] for embed in EmbettyIE._extract_entries(webpage, **kw)]