diff --git a/src/crewai/llm.py b/src/crewai/llm.py
index c8c456297..82e48c8d6 100644
--- a/src/crewai/llm.py
+++ b/src/crewai/llm.py
@@ -338,7 +338,13 @@ class LLM(BaseLLM):
         Returns:
             Dict[str, Any]: Parameters for the completion call
         """
-        # --- 1) Format messages according to provider requirements
+        # --- 1) Ensure messages list is not None or empty (additional safeguard)
+        if messages is None:
+            raise ValueError("Messages list cannot be empty. At least one message is required.")
+        if isinstance(messages, list) and len(messages) == 0:
+            raise ValueError("Messages list cannot be empty. At least one message is required.")
+            
+        # --- 2) Format messages according to provider requirements
         if isinstance(messages, str):
             messages = [{"role": "user", "content": messages}]
         formatted_messages = self._format_messages_for_provider(messages)
@@ -845,7 +851,13 @@ class LLM(BaseLLM):
             ValueError: If response format is not supported
             LLMContextLengthExceededException: If input exceeds model's context limit
         """
-        # --- 1) Emit call started event
+        # --- 1) Validate messages is not None or empty to prevent IndexError in LiteLLM's ollama_pt()
+        if messages is None:
+            raise ValueError("Messages list cannot be empty. At least one message is required.")
+        if isinstance(messages, list) and len(messages) == 0:
+            raise ValueError("Messages list cannot be empty. At least one message is required.")
+
+        # --- 2) Emit call started event
         assert hasattr(crewai_event_bus, "emit")
         crewai_event_bus.emit(
             self,
@@ -857,10 +869,10 @@ class LLM(BaseLLM):
             ),
         )
 
-        # --- 2) Validate parameters before proceeding with the call
+        # --- 3) Validate parameters before proceeding with the call
         self._validate_call_params()
 
-        # --- 3) Convert string messages to proper format if needed
+        # --- 4) Convert string messages to proper format if needed
         if isinstance(messages, str):
             messages = [{"role": "user", "content": messages}]
 
diff --git a/tests/test_empty_messages.py b/tests/test_empty_messages.py
new file mode 100644
index 000000000..96cc5e743
--- /dev/null
+++ b/tests/test_empty_messages.py
@@ -0,0 +1,30 @@
+import pytest
+from unittest.mock import patch
+
+from crewai.llm import LLM
+
+
+def test_empty_messages_validation():
+    """
+    Test that LLM.call() raises a ValueError when an empty messages list is passed.
+    This prevents the IndexError in LiteLLM's ollama_pt() function.
+    """
+    llm = LLM(model="gpt-3.5-turbo")  # Any model will do for this test
+    
+    with pytest.raises(ValueError, match="Messages list cannot be empty"):
+        llm.call(messages=[])
+    
+    with pytest.raises(ValueError, match="Messages list cannot be empty"):
+        llm.call(messages=None)
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_ollama_model_empty_messages():
+    """
+    Test that LLM.call() with an Ollama model raises a ValueError 
+    when an empty messages list is passed.
+    """
+    llm = LLM(model="ollama/llama3")
+    
+    with pytest.raises(ValueError, match="Messages list cannot be empty"):
+        llm.call(messages=[])