diff --git a/servers/fastapi/services/llm_client.py b/servers/fastapi/services/llm_client.py
index e75545a4..a005dd4c 100644
--- a/servers/fastapi/services/llm_client.py
+++ b/servers/fastapi/services/llm_client.py
@@ -79,14 +79,14 @@ class LLMClient:
         )
 
     def _get_custom_client(self):
-        if not (get_custom_llm_api_key_env() and get_custom_llm_url_env()):
+        if not get_custom_llm_url_env():
             raise HTTPException(
                 status_code=400,
-                detail="Custom LLM API Key is not set",
+                detail="Custom LLM URL is not set",
             )
         return AsyncOpenAI(
             base_url=get_custom_llm_url_env(),
-            api_key=get_custom_llm_api_key_env(),
+            api_key=get_custom_llm_api_key_env() or "null",
         )
 
     # ? Prompts
diff --git a/servers/fastapi/utils/model_availability.py b/servers/fastapi/utils/model_availability.py
index a546640e..4f942ac7 100644
--- a/servers/fastapi/utils/model_availability.py
+++ b/servers/fastapi/utils/model_availability.py
@@ -92,15 +92,12 @@ async def check_llm_and_image_provider_api_or_model_availability():
         elif is_custom_llm_selected():
             custom_model = get_custom_model_env()
             custom_llm_url = get_custom_llm_url_env()
-            custom_llm_api_key = get_custom_llm_api_key_env()
             if not custom_model:
                 raise Exception("CUSTOM_MODEL must be provided")
             if not custom_llm_url:
                 raise Exception("CUSTOM_LLM_URL must be provided")
-            if not custom_llm_api_key:
-                raise Exception("CUSTOM_LLM_API_KEY must be provided")
             available_models = await list_available_openai_compatible_models(
-                custom_llm_url, custom_llm_api_key
+                custom_llm_url, get_custom_llm_api_key_env() or "null"
             )
             print("-" * 50)
             print("Available models: ", available_models)