diff --git a/litellm/__init__.py b/litellm/__init__.py
index aaa36737ef..01559e3d18 100644
--- a/litellm/__init__.py
+++ b/litellm/__init__.py
@@ -10,6 +10,8 @@ azure_key = None
 anthropic_key = None 
 replicate_key = None 
 cohere_key = None 
+
+hugging_api_token = None
 ####### THREAD-SPECIFIC DATA ###################
 class MyLocal(threading.local):
     def __init__(self):
diff --git a/litellm/main.py b/litellm/main.py
index 299376458f..4fc3bcb05e 100644
--- a/litellm/main.py
+++ b/litellm/main.py
@@ -44,7 +44,8 @@ def completion(
     temperature=1, top_p=1, n=1, stream=False, stop=None, max_tokens=float('inf'),
     presence_penalty=0, frequency_penalty=0, logit_bias={}, user="", deployment_id=None,
     # Optional liteLLM function params
-    *, return_async=False, api_key=None, force_timeout=60, azure=False, logger_fn=None, verbose=False
+    *, return_async=False, api_key=None, force_timeout=60, azure=False, logger_fn=None, verbose=False,
+    hugging_face = False
   ):
   try:
     global new_response
@@ -273,6 +274,32 @@ def completion(
           "total_tokens": prompt_tokens + completion_tokens
         }
       response = model_response
+    elif hugging_face == True:
+      import requests
+      API_URL = f"https://api-inference.huggingface.co/models/{model}"
+      HF_TOKEN = get_secret("HF_TOKEN")
+      headers = {"Authorization": f"Bearer {HF_TOKEN}"}
+
+      prompt = " ".join([message["content"] for message in messages])
+      ## LOGGING
+      logging(model=model, input=prompt, azure=azure, logger_fn=logger_fn)
+      input_payload = {"inputs": prompt}
+      response = requests.post(API_URL, headers=headers, json=input_payload)
+  
+      completion_response = response.json()[0]['generated_text']
+      ## LOGGING
+      logging(model=model, input=prompt, azure=azure, additional_args={"max_tokens": max_tokens, "original_response": completion_response}, logger_fn=logger_fn)
+      prompt_tokens = len(encoding.encode(prompt))
+      completion_tokens = len(encoding.encode(completion_response))
+      ## RESPONSE OBJECT
+      model_response["choices"][0]["message"]["content"] = completion_response
+      model_response["created"] = time.time()
+      model_response["model"] = model
+      model_response["usage"] = {
+          "prompt_tokens": prompt_tokens,
+          "completion_tokens": completion_tokens,
+          "total_tokens": prompt_tokens + completion_tokens
+        }
     else: 
       ## LOGGING
       logging(model=model, input=messages, azure=azure, logger_fn=logger_fn)
diff --git a/litellm/tests/test_completion.py b/litellm/tests/test_completion.py
index 52c9373cc6..e001daa615 100644
--- a/litellm/tests/test_completion.py
+++ b/litellm/tests/test_completion.py
@@ -26,6 +26,16 @@ def test_completion_claude():
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
 
+def test_completion_hf_api():
+    try:
+        user_message = "write some code to find the sum of two numbers"
+        messages = [{ "content": user_message,"role": "user"}]
+        response = completion(model="stabilityai/stablecode-completion-alpha-3b-4k", messages=messages, hugging_face=True)
+        # Add any assertions here to check the response
+        print(response)
+    except Exception as e:
+        pytest.fail(f"Error occurred: {e}")
+
 def test_completion_cohere():
     try:
         response = completion(model="command-nightly", messages=messages, max_tokens=500)