diff --git a/autoload/leetcode.py b/autoload/leetcode.py
index 436c595..04720d9 100644
--- a/autoload/leetcode.py
+++ b/autoload/leetcode.py
@@ -348,12 +348,12 @@ def _check_result(submission_id):
 
     # the keys differs between the result of testing the code and submitting it
     # for submission judge_type is 'large', and for testing judge_type does not exist
-    if r.get('judge_type') == 'large':
-        result['answer'] = _split(r.get('code_output', ''))
-        result['expected_answer'] = _split(r.get('expected_output', ''))
-        result['stdout'] = _split(r.get('std_output', ''))
-        result['runtime_percentile'] = r.get('runtime_percentile', '')
-    else:
+    # Note: 
+    #   As of October 09 2020, 'judge_type' is not provided in response, 
+    #   however we can use submission_id's value to decide output formatting
+    #   'Run code' Example submission_id: "runcode_1602277499.4352891_sYj3Jte5Lb"   
+    #   'Submit' Example submission_id: "406680787" 
+    if r.get('submission_id').startswith('runcode_'):
         # Test states cannot distinguish accepted answers from wrong answers.
         if result['state'] == 'Accepted':
             result['state'] = 'Finished'
@@ -361,6 +361,11 @@ def _check_result(submission_id):
         result['expected_answer'] = []
         result['runtime_percentile'] = r.get('runtime_percentile', '')
         result['expected_answer'] = r.get('expected_code_answer', [])
+    else:
+        result['answer'] = _split(r.get('code_output', ''))
+        result['expected_answer'] = _split(r.get('expected_output', ''))
+        result['stdout'] = _split(r.get('std_output', ''))
+        result['runtime_percentile'] = r.get('runtime_percentile', '')
     return result