diff --git a/litellm/proxy/litellm_pre_call_utils.py b/litellm/proxy/litellm_pre_call_utils.py
index 9417731cd..3a1c456aa 100644
--- a/litellm/proxy/litellm_pre_call_utils.py
+++ b/litellm/proxy/litellm_pre_call_utils.py
@@ -33,11 +33,11 @@ def _get_metadata_variable_name(request: Request) -> str:
     """
     Helper to return what the "metadata" field should be called in the request data
 
-    For all /thread endpoints we need to call this "litellm_metadata"
+    For all /thread or /assistant endpoints we need to call this "litellm_metadata"
 
     For ALL other endpoints we call this "metadata
     """
-    if "thread" in request.url.path:
+    if "thread" in request.url.path or "assistant" in request.url.path:
         return "litellm_metadata"
     else:
         return "metadata"
diff --git a/litellm/proxy/proxy_server.py b/litellm/proxy/proxy_server.py
index 34352bb8a..65a29e549 100644
--- a/litellm/proxy/proxy_server.py
+++ b/litellm/proxy/proxy_server.py
@@ -3975,10 +3975,10 @@ async def create_assistant(
     API Reference docs - https://platform.openai.com/docs/api-reference/assistants/createAssistant
     """
     global proxy_logging_obj
-    data: Dict = {}
     try:
         # Use orjson to parse JSON data, orjson speeds up requests significantly
         body = await request.body()
+        data = orjson.loads(body)
 
         # Include original request and headers in the data
         data = await add_litellm_data_to_request(