diff --git a/deepdoc/parser/json_parser.py b/deepdoc/parser/json_parser.py
index 249ce1ccb9..36045cbd4c 100644
--- a/deepdoc/parser/json_parser.py
+++ b/deepdoc/parser/json_parser.py
@@ -45,6 +45,15 @@ class RAGFlowJsonParser:
         """Calculate the size of the serialized JSON object."""
         return len(json.dumps(data, ensure_ascii=False))
 
+    @staticmethod
+    def _is_empty_chunk(chunk: Any) -> bool:
+        """Only null and empty containers carry no content; 0 and false do."""
+        if chunk is None:
+            return True
+        if isinstance(chunk, (dict, list, str)):
+            return not chunk
+        return False
+
     @staticmethod
     def _set_nested_dict(d: dict, path: list[str], value: Any) -> None:
         """Set a value in a nested dictionary based on the given path."""
@@ -93,7 +102,11 @@ class RAGFlowJsonParser:
                     self._json_split(value, new_path, chunks)
         else:
             # handle single item
-            self._set_nested_dict(chunks[-1], current_path, data)
+            if not current_path:
+                # top-level scalar (number/string/bool/null) has no key to nest under
+                chunks[-1] = data
+            else:
+                self._set_nested_dict(chunks[-1], current_path, data)
         return chunks
 
     def split_json(
@@ -110,7 +123,7 @@ class RAGFlowJsonParser:
             chunks = self._json_split(json_data, None, None)
 
         # Remove the last chunk if it's empty
-        if not chunks[-1]:
+        if chunks and self._is_empty_chunk(chunks[-1]):
             chunks.pop()
         return chunks
 
@@ -132,7 +145,7 @@ class RAGFlowJsonParser:
         try:
             json_data = json.loads(content)
             chunks = self.split_json(json_data, True)
-            sections = [json.dumps(line, ensure_ascii=False) for line in chunks if line]
+            sections = [json.dumps(line, ensure_ascii=False) for line in chunks if not self._is_empty_chunk(line)]
         except json.JSONDecodeError:
             pass
         return sections
@@ -146,7 +159,7 @@ class RAGFlowJsonParser:
             try:
                 data = json.loads(line)
                 chunks = self.split_json(data, convert_lists=True)
-                all_chunks.extend(json.dumps(chunk, ensure_ascii=False) for chunk in chunks if chunk)
+                all_chunks.extend(json.dumps(chunk, ensure_ascii=False) for chunk in chunks if not self._is_empty_chunk(chunk))
             except json.JSONDecodeError:
                 continue
         return all_chunks
diff --git a/test/unit_test/deepdoc/parser/test_json_parser.py b/test/unit_test/deepdoc/parser/test_json_parser.py
new file mode 100644
index 0000000000..a548bcf452
--- /dev/null
+++ b/test/unit_test/deepdoc/parser/test_json_parser.py
@@ -0,0 +1,86 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+"""Unit tests for RAGFlowJsonParser.
+
+Regression for the case where a .json upload whose top-level value is a bare
+JSON scalar (a number, string, boolean, or null - all valid JSON) reached
+``_json_split`` with an empty ``current_path``. ``_set_nested_dict`` then indexed
+``path[-1]`` on an empty list and raised ``IndexError``, which ``_parse_json``
+does not catch (it only guards ``json.JSONDecodeError``), so the whole upload
+crashed. A top-level scalar has no key to nest under and must be stored as the
+chunk directly.
+"""
+
+import importlib.util
+import os
+import sys
+from unittest import mock
+
+# Load json_parser by file path so we don't trigger deepdoc/parser/__init__.py
+# (which pulls in heavy parsers). json_parser only imports ``find_codec`` from
+# rag.nlp, and only inside ``__call__``; stub rag.nlp so the module imports.
+if "rag" not in sys.modules:
+    sys.modules["rag"] = mock.MagicMock()
+if "rag.nlp" not in sys.modules:
+    sys.modules["rag.nlp"] = mock.MagicMock()
+
+
+def _find_project_root(marker="pyproject.toml"):
+    d = os.path.dirname(os.path.abspath(__file__))
+    while d != os.path.dirname(d):
+        if os.path.exists(os.path.join(d, marker)):
+            return d
+        d = os.path.dirname(d)
+    return None
+
+
+_PROJECT_ROOT = _find_project_root()
+
+_json_spec = importlib.util.spec_from_file_location(
+    "deepdoc.parser.json_parser",
+    os.path.join(_PROJECT_ROOT, "deepdoc", "parser", "json_parser.py"),
+)
+_json_mod = importlib.util.module_from_spec(_json_spec)
+sys.modules["deepdoc.parser.json_parser"] = _json_mod
+_json_spec.loader.exec_module(_json_mod)
+
+RAGFlowJsonParser = _json_mod.RAGFlowJsonParser
+
+
+def test_top_level_scalars_do_not_crash():
+    # Previously raised IndexError instead of returning a chunk.
+    parser = RAGFlowJsonParser()
+    assert parser._parse_json("42") == ["42"]
+    assert parser._parse_json('"hello"') == ['"hello"']
+    assert parser._parse_json("true") == ["true"]
+    assert parser._parse_json("0") == ["0"]
+    assert parser._parse_json("false") == ["false"]
+
+
+def test_top_level_null_yields_no_chunk():
+    # null carries no content; it should be dropped, not crash.
+    parser = RAGFlowJsonParser()
+    assert parser._parse_json("null") == []
+    assert parser._parse_json('""') == []
+    assert parser._parse_json("{}") == []
+    assert parser._parse_json("[]") == []
+
+
+def test_objects_and_arrays_still_chunk():
+    parser = RAGFlowJsonParser()
+    assert parser._parse_json('{"a": 1}') == ['{"a": 1}']
+    assert parser._parse_json("[1, 2, 3]") != []