diff --git a/litellm/llms/prompt_templates/factory.py b/litellm/llms/prompt_templates/factory.py
index f81515e98..7ec21e8bb 100644
--- a/litellm/llms/prompt_templates/factory.py
+++ b/litellm/llms/prompt_templates/factory.py
@@ -1010,6 +1010,9 @@ def convert_to_gemini_tool_call_invoke(
                 name = tool["function"].get("name", "")
                 arguments = tool["function"].get("arguments", "")
                 arguments_dict = json.loads(arguments)
+                function_call: Optional[litellm.types.llms.vertex_ai.FunctionCall] = (
+                    None
+                )
                 for k, v in arguments_dict.items():
                     inferred_protocol_value = infer_protocol_value(value=v)
                     _field = litellm.types.llms.vertex_ai.Field(
@@ -1022,9 +1025,18 @@ def convert_to_gemini_tool_call_invoke(
                         name=name,
                         args=_fields,
                     )
-                _parts_list.append(
-                    litellm.types.llms.vertex_ai.PartType(function_call=function_call)
-                )
+                if function_call is not None:
+                    _parts_list.append(
+                        litellm.types.llms.vertex_ai.PartType(
+                            function_call=function_call
+                        )
+                    )
+                else:  # don't silently drop params. Make it clear to user what's happening.
+                    raise Exception(
+                        "function_call missing. Received tool call with 'type': 'function'. No function call in argument - {}".format(
+                            tool
+                        )
+                    )
         return _parts_list
     except Exception as e:
         raise Exception(
diff --git a/litellm/llms/vertex_httpx.py b/litellm/llms/vertex_httpx.py
index b92730613..8af8f6894 100644
--- a/litellm/llms/vertex_httpx.py
+++ b/litellm/llms/vertex_httpx.py
@@ -491,6 +491,16 @@ class VertexGeminiConfig:
             "SPII": "The token generation was stopped as the response was flagged for Sensitive Personally Identifiable Information (SPII) contents.",
         }
 
+    def translate_exception_str(self, exception_string: str):
+        if (
+            "GenerateContentRequest.tools[0].function_declarations[0].parameters.properties: should be non-empty for OBJECT type"
+            in exception_string
+        ):
+            return "'properties' field in tools[0]['function']['parameters'] cannot be empty if 'type' == 'object'. Received error from provider - {}".format(
+                exception_string
+            )
+        return exception_string
+
 
 async def make_call(
     client: Optional[AsyncHTTPHandler],
@@ -504,8 +514,15 @@ async def make_call(
     if client is None:
         client = AsyncHTTPHandler()  # Create a new client if none provided
 
-    response = await client.post(api_base, headers=headers, data=data, stream=True)
-
+    try:
+        response = await client.post(api_base, headers=headers, data=data, stream=True)
+        response.raise_for_status()
+    except httpx.HTTPStatusError as e:
+        exception_string = str(await e.response.aread())
+        raise VertexAIError(
+            status_code=e.response.status_code,
+            message=VertexGeminiConfig().translate_exception_str(exception_string),
+        )
     if response.status_code != 200:
         raise VertexAIError(status_code=response.status_code, message=response.text)
 
diff --git a/litellm/proxy/pass_through_endpoints/pass_through_endpoints.py b/litellm/proxy/pass_through_endpoints/pass_through_endpoints.py
index 511ce1ee9..966593551 100644
--- a/litellm/proxy/pass_through_endpoints/pass_through_endpoints.py
+++ b/litellm/proxy/pass_through_endpoints/pass_through_endpoints.py
@@ -384,8 +384,16 @@ async def pass_through_request(
                 params=requested_query_params,
                 headers=headers,
             )
+
             response = await async_client.send(req, stream=stream)
 
+            try:
+                response.raise_for_status()
+            except httpx.HTTPStatusError as e:
+                raise HTTPException(
+                    status_code=e.response.status_code, detail=await e.response.aread()
+                )
+
             # Create an async generator to yield the response content
             async def stream_response() -> AsyncIterable[bytes]:
                 async for chunk in response.aiter_bytes():
@@ -409,6 +417,13 @@ async def pass_through_request(
             response.headers.get("content-type") is not None
             and response.headers["content-type"] == "text/event-stream"
         ):
+            try:
+                response.raise_for_status()
+            except httpx.HTTPStatusError as e:
+                raise HTTPException(
+                    status_code=e.response.status_code, detail=await e.response.aread()
+                )
+
             # streaming response
             # Create an async generator to yield the response content
             async def stream_response() -> AsyncIterable[bytes]:
@@ -421,6 +436,13 @@ async def pass_through_request(
                 status_code=response.status_code,
             )
 
+        try:
+            response.raise_for_status()
+        except httpx.HTTPStatusError as e:
+            raise HTTPException(
+                status_code=e.response.status_code, detail=e.response.text
+            )
+
         if response.status_code >= 300:
             raise HTTPException(status_code=response.status_code, detail=response.text)