diff --git a/litellm/__pycache__/main.cpython-311.pyc b/litellm/__pycache__/main.cpython-311.pyc
index 42ce552ba..521efeb32 100644
Binary files a/litellm/__pycache__/main.cpython-311.pyc and b/litellm/__pycache__/main.cpython-311.pyc differ
diff --git a/litellm/__pycache__/utils.cpython-311.pyc b/litellm/__pycache__/utils.cpython-311.pyc
index 4c39a1c48..af8acfef8 100644
Binary files a/litellm/__pycache__/utils.cpython-311.pyc and b/litellm/__pycache__/utils.cpython-311.pyc differ
diff --git a/litellm/main.py b/litellm/main.py
index b02e8cebd..3f37ace14 100644
--- a/litellm/main.py
+++ b/litellm/main.py
@@ -257,7 +257,7 @@ def completion(
         if deployment_id != None: # azure llms 
                 model=deployment_id
                 custom_llm_provider="azure"
-        model, custom_llm_provider = get_llm_provider(model=model, custom_llm_provider=custom_llm_provider, api_base=api_base)
+        model, custom_llm_provider, dynamic_api_key = get_llm_provider(model=model, custom_llm_provider=custom_llm_provider, api_base=api_base)
         model_api_key = get_api_key(llm_provider=custom_llm_provider, dynamic_api_key=api_key) # get the api key from the environment if required for the model
         if model_api_key and "sk-litellm" in model_api_key:
             api_base = "https://proxy.litellm.ai"
@@ -391,6 +391,7 @@ def completion(
             # set API KEY
             api_key = (
                 api_key or
+                dynamic_api_key or # allows us to read env variables for compatible openai api's like perplexity 
                 litellm.api_key or
                 litellm.openai_key or
                 get_secret("OPENAI_API_KEY")
@@ -1371,7 +1372,7 @@ def embedding(
     caching=False,
     custom_llm_provider=None,
 ):
-    model, custom_llm_provider = get_llm_provider(model, custom_llm_provider)
+    model, custom_llm_provider, dynamic_api_key = get_llm_provider(model, custom_llm_provider)
     try:
         response = None
         logging = litellm_logging_obj
diff --git a/litellm/utils.py b/litellm/utils.py
index 74e81fa8f..08d9614f8 100644
--- a/litellm/utils.py
+++ b/litellm/utils.py
@@ -1405,15 +1405,16 @@ def get_optional_params(  # use the openai defaults
 
 def get_llm_provider(model: str, custom_llm_provider: Optional[str] = None, api_base: Optional[str] = None):
     try:
+        dynamic_api_key = None
         # check if llm provider provided
         if custom_llm_provider:
-            return model, custom_llm_provider
+            return model, custom_llm_provider, dynamic_api_key
 
         # check if llm provider part of model name
         if model.split("/",1)[0] in litellm.provider_list:
             custom_llm_provider = model.split("/", 1)[0]
             model = model.split("/", 1)[1]
-            return model, custom_llm_provider
+            return model, custom_llm_provider, dynamic_api_key
 
         # check if api base is a known openai compatible endpoint
         if api_base: 
@@ -1421,8 +1422,8 @@ def get_llm_provider(model: str, custom_llm_provider: Optional[str] = None, api_
                 if endpoint in api_base:
                     custom_llm_provider = "custom_openai"
                     if endpoint == "api.perplexity.ai": 
-                        litellm.api_key = os.getenv("PERPLEXITYAI_API_KEY")
-                    return model, custom_llm_provider
+                        dynamic_api_key = os.getenv("PERPLEXITYAI_API_KEY")
+                    return model, custom_llm_provider, dynamic_api_key
 
         # check if model in known model provider list  -> for huggingface models, raise exception as they don't have a fixed provider (can be togetherai, anyscale, baseten, runpod, et.)
         ## openai - chatcompletion + text completion
@@ -1479,7 +1480,7 @@ def get_llm_provider(model: str, custom_llm_provider: Optional[str] = None, api_
             print("\033[1;31mProvider List: https://docs.litellm.ai/docs/providers\033[0m")
             print()
             raise ValueError(f"LLM Provider NOT provided. Pass in the LLM provider you are trying to call. E.g. For 'Huggingface' inference endpoints pass in `completion(model='huggingface/{model}',..)` Learn more: https://docs.litellm.ai/docs/providers")
-        return model, custom_llm_provider
+        return model, custom_llm_provider, dynamic_api_key
     except Exception as e: 
         raise e