diff --git a/src/parsing/utils.py b/src/parsing/utils.py
index edd6cdbc2e..65abede732 100644
--- a/src/parsing/utils.py
+++ b/src/parsing/utils.py
@@ -258,8 +258,8 @@ class EntryParsed:
         enclosures: list[Enclosure] = None
 
     content = (
-            entry.get('content')  # Atom
-            or entry.get('summary', '')  # Atom summary or RSS description
+            entry.get('content')  # Atom: <content>; JSON Feed: .content_html, .content_text
+            or entry.get('summary', '')  # Atom: <summary>; RSS: <description>
     )
 
     if isinstance(content, list) and len(content) > 0:  # Atom
@@ -271,6 +271,9 @@ class EntryParsed:
         else:
             content = content[0]
         content = content.get('value', '')
+    elif isinstance(content, dict):  # JSON Feed
+        # TODO: currently feedparser always prefer content_text rather than content_html, we'd like to change that
+        content = content.get('value', '')
 
     EntryParsed.content = await html_validator(content)
     EntryParsed.link = entry.get('link') or entry.get('guid')