fix: fix response class strictness

packages/gen/gen_ai_hub/orchestration_v2/models/base.py

@@ -31,3 +31,16 @@ def model_dump(self, **kwargs):
         kwargs.setdefault("by_alias", True)
         kwargs.setdefault("exclude_none", True)
         return super().model_dump(**kwargs)
+
+class ResponseBaseModel(BaseModel):
+    """
+    Base model for API response models.
+
+    - `extra="allow"` allows unexpected fields in responses to be accepted,
+      since the external API might introduce new attributes in the response.
+    """
+
+    model_config = ConfigDict(
+        extra="allow",
+        frozen=False,
+    )

packages/gen/gen_ai_hub/orchestration_v2/models/embeddings.py

@@ -7,7 +7,7 @@
 
 from pydantic import Field
 
-from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel
+from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel, ResponseBaseModel
 from gen_ai_hub.orchestration_v2.models.data_masking import MaskingModuleConfig
 
 
@@ -117,7 +117,7 @@ class EmbeddingsInput(BaseModel):
     type_: Optional[EmbeddingsInputType] = Field(default=None, alias="type")
 
 
-class EmbeddingsUsage(BaseModel):
+class EmbeddingsUsage(ResponseBaseModel):
     """
     Token usage information for the embeddings request.
 
@@ -129,7 +129,7 @@ class EmbeddingsUsage(BaseModel):
     total_tokens: int
 
 
-class EmbeddingResult(BaseModel):
+class EmbeddingResult(ResponseBaseModel):
     """
     A single embedding result.
 
@@ -143,7 +143,7 @@ class EmbeddingResult(BaseModel):
     index: int
 
 
-class EmbeddingsResponse(BaseModel):
+class EmbeddingsResponse(ResponseBaseModel):
     """
     The response from the embedding model, following OpenAI specification.
 
@@ -159,7 +159,7 @@ class EmbeddingsResponse(BaseModel):
     usage: EmbeddingsUsage
 
 
-class EmbeddingsPostResponse(BaseModel):
+class EmbeddingsPostResponse(ResponseBaseModel):
     """
     Response for an embeddings POST request.
 

packages/gen/gen_ai_hub/orchestration_v2/models/message.py

@@ -10,11 +10,11 @@
 
 from pydantic import field_validator, ValidationError
 
-from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel
+from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel, ResponseBaseModel
 from gen_ai_hub.orchestration_v2.models.multimodal_items import ContentPart, ImageItem, TextPart, ImageUrl, ImagePart
 
 
-class FunctionCall(BaseModel):
+class FunctionCall(ResponseBaseModel):
     """
     Represents a function call with its name and arguments.
 
@@ -44,7 +44,7 @@ def parse_arguments(self) -> dict:
         return json.loads(self.arguments)
 
 
-class MessageToolCall(BaseModel):
+class MessageToolCall(ResponseBaseModel):
     """
     The tool calls generated by the model, such as function calls.
 
@@ -71,6 +71,7 @@ class ReasoningBlock(BaseModel):
     signature: str
 
 
+
 class Role(str, Enum):
     """
     Enumerates supported roles in LLM-based conversations.
@@ -180,7 +181,7 @@ class DeveloperChatMessage(BaseModel):
     role: Role = Role.DEVELOPER
     content: Union[str, List[TextPart]]
 
-class ResponseChatMessage(BaseModel):
+class ResponseChatMessage(ResponseBaseModel):
     """
     Represents a response message in a conversation.
 

packages/gen/gen_ai_hub/orchestration_v2/models/response.py

@@ -6,26 +6,10 @@
 from pydantic import ConfigDict, Field
 
 from gen_ai_hub.orchestration.models.response import ModuleResultsStreaming
-from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel
+from gen_ai_hub.orchestration_v2.models.base import ResponseBaseModel
 from gen_ai_hub.orchestration_v2.models.message import ChatMessage, FunctionCall, ResponseChatMessage
 
 
-class ResponseBaseModel(BaseModel):
-    """
-    Abstract base model that extends Pydantic's BaseModel and ABC.
-
-    - `extra="allow"` allows unexpected fields in responses to be accepted,
-      since the external API might introduce new attributes in the response.
-
-    This enforces consistent and safe serialization behavior across all
-    derived models.
-    """
-
-    model_config = ConfigDict(
-        extra="allow",
-        frozen=False,
-    )
-
 
 class CacheCreationTokenDetails(ResponseBaseModel):
     """

packages/gen/tests/orchestration_v2/test_embeddings.py

@@ -734,6 +734,47 @@ def test_request_with_masking_allowlist(self):
         allowlist = result["config"]["modules"]["masking"]["masking_providers"][0]["allowlist"]
         self.assertEqual(allowlist, ["SAP", "Microsoft"])
 
+class TestEmbeddingsResponseExtraFields(unittest.TestCase):
+    """EmbeddingsUsage, EmbeddingResult, EmbeddingsResponse and EmbeddingsPostResponse
+    were switched to ResponseBaseModel and must silently accept unknown fields."""
+
+    def test_embeddings_usage_stores_extra_field(self):
+        usage = EmbeddingsUsage.model_validate({
+            "prompt_tokens": 10, "total_tokens": 10,
+            "extra_field": "extra",
+        })
+        self.assertEqual(usage.extra_field, "extra")
+
+    def test_embedding_result_stores_extra_field(self):
+        result = EmbeddingResult.model_validate({
+            "object": "embedding", "embedding": [0.1, 0.2], "index": 0,
+            "extra_field": "extra",
+        })
+        self.assertEqual(result.extra_field, "extra")
+
+    def test_embeddings_response_stores_extra_field(self):
+        response = EmbeddingsResponse.model_validate({
+            "object": "list",
+            "data": [{"object": "embedding", "embedding": [0.1], "index": 0}],
+            "model": "text-embedding-3-large",
+            "usage": {"prompt_tokens": 5, "total_tokens": 5},
+            "extra_field": "extra",
+        })
+        self.assertEqual(response.extra_field, "extra")
+
+    def test_embeddings_post_response_stores_extra_field(self):
+        response = EmbeddingsPostResponse.model_validate({
+            "request_id": "emb-req-1",
+            "final_result": {
+                "object": "list",
+                "data": [{"object": "embedding", "embedding": [0.1], "index": 0}],
+                "model": "text-embedding-3-large",
+                "usage": {"prompt_tokens": 5, "total_tokens": 5},
+            },
+            "extra_field": "extra",
+        })
+        self.assertEqual(response.extra_field, "extra")
+
 
 if __name__ == "__main__":
     unittest.main()