fix(server): update /v1/chat/completions endpoint to be backwards compatabile

Author: christinaexyou

nemoguardrails/server/schemas/openai.py

@@ -54,6 +54,16 @@ class OpenAIChatCompletionRequest(BaseModel):
         ...,
         description="The LLM model to use for chat completion (e.g., 'gpt-4o', 'llama-3.1-8b').",
     )
+
+    @field_validator("model", mode="before")
+    @classmethod
+    def normalize_model(cls, v: Any) -> Optional[str]:
+        if isinstance(v, dict):
+            if not isinstance(v.get("id"), str):
+                raise ValueError("Model object must contain a string 'id' field")
+            return v["id"]
+        return v
+
     stream: Optional[bool] = Field(
         default=False,
         description="If set, partial message deltas will be sent as server-sent events.",

tests/server/test_openai_integration.py

@@ -86,6 +86,23 @@ def test_openai_client_chat_completion(openai_client):
     assert hasattr(response, "created")
 
 
+def test_chat_completion_model_as_dict():
+    test_client = TestClient(api.app)
+    response = test_client.post(
+        "/v1/chat/completions",
+        json={
+            "model": {"id": "gpt-4o", "name": "gpt-4o", "maxLength": 10000},
+            "messages": [{"role": "user", "content": "hi"}],
+            "guardrails": {"config_id": "with_custom_llm"},
+        },
+    )
+    assert response.status_code == 200
+    data = response.json()
+    assert data["model"] == "gpt-4o"
+    assert data["choices"][0]["message"]["content"] == "Custom LLM response"
+    assert data["choices"][0]["finish_reason"] == "stop"
+
+
 def test_openai_client_chat_completion_parameterized(openai_client):
     response = openai_client.chat.completions.create(
         model="gpt-4o",