diff --git a/openapi_python_client/parser/openapi.py b/openapi_python_client/parser/openapi.py
index 4f83ae93e..94069eee3 100644
--- a/openapi_python_client/parser/openapi.py
+++ b/openapi_python_client/parser/openapi.py
@@ -25,7 +25,7 @@
     property_from_data,
 )
 from .properties.schemas import parameter_from_reference
-from .responses import HTTPStatusPattern, Responses, response_from_data
+from .responses import HTTPStatusPattern, JSON_SOURCE, Responses, response_from_data, source_by_content_type
 
 _PATH_PARAM_REGEX = re.compile("{([a-zA-Z_-][a-zA-Z0-9_-]*)}")
 
@@ -250,7 +250,20 @@ def add_parameters(
             param = param_or_error  # noqa: PLW2901
 
             if param.param_schema is None:
-                continue
+                if param.content:
+                    content_schemas = []
+                    for content_type in param.content.keys():
+                        if ((content := param.content.get(content_type)) is not None
+                            and content.media_type_schema is not None
+                            and source_by_content_type(content_type, config) is JSON_SOURCE
+                        ):
+                            content_schemas.append(content.media_type_schema)
+                    if len(content_schemas) == 1:
+                        param.param_schema = content_schemas[0]
+                    elif len(content_schemas) > 1:
+                        param.param_schema = oai.Schema(oneOf=content_schemas)
+                else:
+                    continue
 
             unique_param = (param.name, param.param_in)
             if unique_param in unique_parameters:
diff --git a/openapi_python_client/parser/responses.py b/openapi_python_client/parser/responses.py
index f05fbce75..6788e68cf 100644
--- a/openapi_python_client/parser/responses.py
+++ b/openapi_python_client/parser/responses.py
@@ -1,4 +1,14 @@
-__all__ = ["HTTPStatusPattern", "Response", "Responses", "response_from_data"]
+__all__ = [
+    "BYTES_SOURCE",
+    "HTTPStatusPattern",
+    "JSON_SOURCE",
+    "NONE_SOURCE",
+    "Response",
+    "response_from_data",
+    "Responses",
+    "source_by_content_type",
+    "TEXT_SOURCE",
+]
 
 from collections.abc import Iterator
 from typing import TypedDict
@@ -128,7 +138,7 @@ def __lt__(self, other: "Response") -> bool:
         return self.status_code < other.status_code
 
 
-def _source_by_content_type(content_type: str, config: Config) -> _ResponseSource | None:
+def source_by_content_type(content_type: str, config: Config) -> _ResponseSource | None:
     parsed_content_type = utils.get_content_type(content_type, config)
     if parsed_content_type is None:
         return None
@@ -207,18 +217,35 @@ def response_from_data(  # noqa: PLR0911
             schemas,
         )
 
+    schemas_ = []
+    source = None
+
     for content_type, media_type in content.items():
-        source = _source_by_content_type(content_type, config)
-        if source is not None:
-            schema_data = media_type.media_type_schema
-            break
-    else:
+        source_ = source_by_content_type(content_type, config)
+
+        if source_ is None:
+            continue
+
+        if source is not None and source_ != source:
+            return (
+                ParseError(
+                    data=data,
+                    detail="Multiple response content types with different attributes are not supported",
+                ),
+                schemas,
+            )
+        source = source_
+
+        if media_type.media_type_schema is not None:
+            schemas_.append(media_type.media_type_schema)
+
+    if source is None:
         return (
-            ParseError(data=data, detail=f"Unsupported content_type {content}"),
+            ParseError(data=data, detail=f"Unsupported content_type(s) {content}"),
             schemas,
         )
 
-    if schema_data is None:
+    if not schemas_:
         return (
             empty_response(
                 status_code=status_code,
@@ -229,6 +256,11 @@ def response_from_data(  # noqa: PLR0911
             schemas,
         )
 
+    if len(schemas_) == 1:
+        schema_data = schemas_[0]
+    else:
+        schema_data = oai.Schema(oneOf=schemas_)
+
     prop, schemas = property_from_data(
         name=response_name,
         required=True,