diff --git a/src/crewai/llm.py b/src/crewai/llm.py
index 01543c14c..00b0ebbe3 100644
--- a/src/crewai/llm.py
+++ b/src/crewai/llm.py
@@ -842,19 +842,44 @@ class LLM:
             return "azure"
             
         return None
+        
+    def _validate_azure_credentials(self) -> bool:
+        """
+        Validates that all required Azure credentials are present and of the correct type.
+        
+        Returns:
+            bool: True if all Azure credentials are valid, False otherwise.
+        """
+        return bool(
+            self.api_key 
+            and self.api_base 
+            and self.api_version 
+            and all(isinstance(x, str) for x in [self.api_key, self.api_base, self.api_version])
+        )
 
     def _validate_call_params(self) -> None:
         """
-        Validate parameters before making a call. Currently this only checks if
-        a response_format is provided and whether the model supports it.
+        Validate parameters before making a call.
+        
         The custom_llm_provider is dynamically determined from the model:
           - E.g., "openrouter/deepseek/deepseek-chat" yields "openrouter"
           - "gemini/gemini-1.5-pro" yields "gemini"
           - "azure/gpt-4" yields "azure"
           - If Azure parameters (api_key, api_base, api_version) are present, "azure" is used
           - If no slash is present and no Azure parameters, None is returned
+          
+        Raises:
+            ValueError: If response_format is not supported by the model or if Azure credentials are incomplete.
         """
         provider = self._get_custom_llm_provider()
+        
+        # Validate Azure credentials if provider is Azure
+        if provider == "azure" and not self._validate_azure_credentials():
+            raise ValueError(
+                "Incomplete Azure credentials. Please provide api_key, api_base, and api_version as strings."
+            )
+            
+        # Validate response_format
         if self.response_format is not None and not supports_response_schema(
             model=self.model,
             custom_llm_provider=provider,
diff --git a/tests/integration/test_azure_credentials.py b/tests/integration/test_azure_credentials.py
new file mode 100644
index 000000000..fa3d30d7c
--- /dev/null
+++ b/tests/integration/test_azure_credentials.py
@@ -0,0 +1,38 @@
+import pytest
+from src.crewai.llm import LLM
+
+def test_azure_detection_with_credentials():
+    """Test that Azure is detected correctly when credentials are provided but model lacks azure/ prefix."""
+    # Create LLM instance with Azure parameters but without azure/ prefix
+    llm = LLM(
+        api_key='test_key',
+        api_base='test_base',
+        model='gpt-4o-mini-2024-07-18',  # Model from issue #2358
+        api_version='test_version'
+    )
+    
+    # Check if provider is detected correctly
+    provider = llm._get_custom_llm_provider()
+    assert provider == 'azure', "Azure provider should be detected based on credentials"
+    
+    # Prepare parameters that would be passed to LiteLLM
+    params = llm._prepare_completion_params(messages=[{"role": "user", "content": "test"}])
+    assert params.get('api_key') == 'test_key', "API key should be included in params"
+    assert params.get('api_base') == 'test_base', "API base should be included in params"
+    assert params.get('api_version') == 'test_version', "API version should be included in params"
+
+
+def test_azure_validation_error():
+    """Test that validation error is raised when Azure credentials are incomplete."""
+    # Create LLM instance with incomplete Azure parameters
+    llm = LLM(
+        model='azure/gpt-4',
+        api_key='test_key',
+        # Missing api_base and api_version
+    )
+    
+    # Validation should fail
+    with pytest.raises(ValueError) as excinfo:
+        llm._validate_call_params()
+    
+    assert "Incomplete Azure credentials" in str(excinfo.value)
diff --git a/tests/llm_test.py b/tests/llm_test.py
index 32c4830aa..23795b8d5 100644
--- a/tests/llm_test.py
+++ b/tests/llm_test.py
@@ -267,6 +267,51 @@ def test_validate_call_params_no_response_format():
     llm._validate_call_params()
 
 
+def test_validate_azure_credentials_valid():
+    llm = LLM(
+        model="gpt-4",
+        api_key="test_key",
+        api_base="test_base",
+        api_version="test_version"
+    )
+    assert llm._validate_azure_credentials() == True
+
+
+def test_validate_azure_credentials_invalid():
+    # Missing api_version
+    llm = LLM(
+        model="gpt-4",
+        api_key="test_key",
+        api_base="test_base"
+    )
+    assert llm._validate_azure_credentials() == False
+    
+    # Non-string value
+    llm = LLM(
+        model="gpt-4",
+        api_key="test_key",
+        api_base="test_base",
+        api_version=123  # Not a string
+    )
+    assert llm._validate_azure_credentials() == False
+
+
+def test_validate_call_params_azure_invalid():
+    # Test with incomplete Azure credentials
+    llm = LLM(
+        model="azure/gpt-4",
+        api_key="test_key",
+        # Missing api_base and api_version
+    )
+    
+    # Should raise ValueError due to incomplete credentials
+    with pytest.raises(ValueError) as excinfo:
+        llm._validate_call_params()
+    
+    # Check error message
+    assert "Incomplete Azure credentials" in str(excinfo.value)
+
+
 @pytest.mark.vcr(filter_headers=["authorization"])
 def test_o3_mini_reasoning_effort_high():
     llm = LLM(