Pass model instead of model_id to AnyLLM

src/django_ai_core/llm/base.py

@@ -21,12 +21,10 @@ def service_id(self) -> str:
         return f"{self.__class__.__name__}:{self.client.PROVIDER_NAME}:{self.model}"
 
     def completion(self, messages, **kwargs):
-        return self.client.completion(model_id=self.model, messages=messages, **kwargs)
+        return self.client.completion(model=self.model, messages=messages, **kwargs)
 
     def responses(self, input_data, **kwargs):
-        return self.client.responses(
-            model_id=self.model, input_data=input_data, **kwargs
-        )
+        return self.client.responses(model=self.model, input_data=input_data, **kwargs)
 
     def embedding(self, inputs, **kwargs):
-        return self.client._embedding(model_id=self.model, inputs=inputs, **kwargs)
+        return self.client._embedding(model=self.model, inputs=inputs, **kwargs)

tests/testapp/indexes.py

@@ -15,13 +15,13 @@
 
 
 class MockAnyLLM(AnyLLM):
-    def completion(self, *, model_id, messages):
+    def completion(self, *, model, messages):
         return "completion"
 
-    def responses(self, *, model_id, input_data):
+    def responses(self, *, model, input_data):
         return "responses"
 
-    def _embedding(self, *, model_id, inputs):
+    def _embedding(self, *, model, inputs):
         return [0, 1, 2]
 
 

tests/unit/llm/test_llm_service.py

@@ -25,9 +25,8 @@ def test_llm_service_completion_wraps_anyllm(mock_any_llm):
     ]
     service = LLMService(client=mock_any_llm, model="mock-model")
     service.completion(messages)
-    print(mock_any_llm.completion.call_args_list)
     mock_any_llm.completion.assert_called_once_with(
-        model_id="mock-model", messages=messages
+        model="mock-model", messages=messages
     )
 
 
@@ -36,14 +35,12 @@ def test_llm_service_responses_wraps_anyllm(mock_any_llm):
     service = LLMService(client=mock_any_llm, model="mock-model")
     service.responses(prompt)
     mock_any_llm.responses.assert_called_once_with(
-        model_id="mock-model", input_data=prompt
+        model="mock-model", input_data=prompt
     )
 
 
 def test_llm_service_embedding_wraps_anyllm(mock_any_llm):
     prompt = "What is the airspeed velocity of an unladen swallow?"
     service = LLMService(client=mock_any_llm, model="mock-model")
     service.embedding(prompt)
-    mock_any_llm._embedding.assert_called_once_with(
-        model_id="mock-model", inputs=prompt
-    )
+    mock_any_llm._embedding.assert_called_once_with(model="mock-model", inputs=prompt)