diff --git a/test/test_InfoExtractor.py b/test/test_InfoExtractor.py
index a16bc16ef..d22b61f62 100644
--- a/test/test_InfoExtractor.py
+++ b/test/test_InfoExtractor.py
@@ -230,7 +230,7 @@ def test_search_json_ld_realworld(self):
                     'description': 'md5:16756d0a18f33bf550e683d134a72f3c',
                     'modified_timestamp': 1636523573,
                     'release_timestamp': 1636523400,
-                    'tags': ['μαχαίρωμα,συμμορία ανηλίκων,ΕΙΔΗΣΕΙΣ,ΕΙΔΗΣΕΙΣ ΣΗΜΕΡΑ,ΝΕΑ,Κοινωνία - Ant1news'],
+                    'tags': 'count:6',
                     'thumbnails': [{'url': 'https://ant1media.azureedge.net/imgHandler/1100/a635c968-be71-447c-bf9c-80d843ece21e.jpg'}],
                     'uploader': 'Ant1news',
                 },
@@ -376,7 +376,7 @@ def test_search_json_ld_realworld(self):
                     'description': 'md5:40eaf402631e0a77d8d74f66574bb978',
                     'modified_timestamp': 1747319520,
                     'release_timestamp': 1747319520,
-                    'tags': ['enseignement secondaire'],
+                    'tags': 'count:1',
                     'timestamp': 1747319520,
                     'thumbnails': [{'url': 'https://www.telemb.be/cdn/ff/pKwkkhB7a5GqSf98QdDUcn9WlvGTYyilvXisHO3fHpI/1747320854/public/2025-05/00006554_avc-tmb-093031.jpeg'}],
                     'uploader': 'Tele MB',
diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py
index d3abba036..74ed84050 100644
--- a/yt_dlp/extractor/common.py
+++ b/yt_dlp/extractor/common.py
@@ -1748,7 +1748,7 @@ def traverse_json_ld(json_ld, at_top_level=True):
                         'description': (('description', 'articleBody'), {str}, {unescapeHTML}, any),
                         'modified_timestamp': ('dateModified', {parse_iso8601}),
                         'release_timestamp': ('datePublished', {parse_iso8601}),
-                        'tags': ('keywords', {str}, {unescapeHTML}, filter, all),
+                        'tags': ('keywords', {str}, {unescapeHTML}, {lambda x: x.split(',')}, filter),
                         'thumbnails': ('image', ..., {
                             'url': ({str}, {unescapeHTML}, {self._proto_relative_url}, {url_or_none}),
                         }),