fix(images): bill proxy image edits at the deployment's own prices

The proxy builds its logging object before the router picks a deployment,
so the deployment's pricing params only reach image_edit as kwargs. Its
pre-call logging update merged the edit request params but not those
kwargs, so custom pricing stayed off and every proxy image edit billed
catalog prices. Image generations already honor them

Merge the call's custom pricing params into the logging litellm_params,
the way chat and generations get them

litellm/images/main.py

@@ -51,6 +51,7 @@ from litellm.types.llms.openai import ImageGenerationRequestQuality
 from litellm.types.router import GenericLiteLLMParams
 from litellm.types.utils import (
     LITELLM_IMAGE_VARIATION_PROVIDERS,
+    CustomPricingLiteLLMParams,
     LlmProviders,
     all_litellm_params,
 )
@@ -883,6 +884,7 @@ def image_edit(
             optional_params=dict(image_edit_request_params),
             litellm_params={
                 **image_edit_request_params,
+                **litellm_params.model_dump(include=set(CustomPricingLiteLLMParams.model_fields), exclude_none=True),
                 "litellm_call_id": litellm_call_id,
                 "model_info": model_info,
             },

tests/unit/llms/azure_ai/image_edit/test_azure_ai_image_edit_transformation.py

@@ -6,6 +6,7 @@ import pathlib
 import struct
 import tempfile
 from collections.abc import Callable, Iterator, Mapping
+from datetime import datetime
 from typing import Final
 
 import httpx
@@ -564,6 +565,52 @@ async def test_flux2_router_image_edit_bills_the_deployment_rates(monkeypatch: p
     )
 
 
+@pytest.mark.parametrize(
+    ("deployment_prices", "expected_cost"),
+    (
+        ({"output_cost_per_image": 0.5}, 0.5),
+        (
+            {"input_cost_per_pixel": 1e-07, "input_cost_per_reference_pixel": 2e-07},
+            1e-07 * 2 * 1024 * 1024 + 2e-07 * 2 * 1024 * 1024,
+        ),
+    ),
+    ids=("flat", "per-pixel"),
+)
+async def test_flux2_router_image_edit_bills_the_deployment_rates_with_a_logger_built_before_routing(
+    monkeypatch: pytest.MonkeyPatch, deployment_prices: Mapping[str, float], expected_cost: float
+):
+    mock_client: Final = AsyncHTTPHandler()
+    mock_client.client = httpx.AsyncClient(transport=httpx.MockTransport(_edit_ok))
+    monkeypatch.setattr(llm_http_handler_module, "get_async_httpx_client", lambda **_kwargs: mock_client)
+    router: Final = litellm.Router(
+        model_list=[
+            {
+                "model_name": "flux2-pro-deployment",
+                "litellm_params": {
+                    "model": "azure_ai/flux.2-pro",
+                    "api_base": "https://example.services.ai.azure.com",
+                    "api_key": "test-key",
+                    **deployment_prices,
+                },
+            }
+        ]
+    )
+    request: Final = {
+        "model": "flux2-pro-deployment",
+        "prompt": "Make it a watercolor",
+        "image": [_png(1024, 1280)],
+        "size": "1024x1280",
+        "litellm_call_id": "proxy-call-id",
+    }
+    logging_obj, routed_request = litellm.utils.function_setup(
+        original_function="aimage_edit", rules_obj=litellm.utils.Rules(), start_time=datetime.now(), **request
+    )
+
+    response: Final = await router.aimage_edit(**routed_request, litellm_logging_obj=logging_obj)
+
+    assert response._hidden_params["response_cost"] == pytest.approx(expected_cost)
+
+
 def _edit_returning(image: bytes) -> Callable[[httpx.Request], httpx.Response]:
     def respond(request: httpx.Request) -> httpx.Response:
         return httpx.Response(200, json={"data": [{"b64_json": base64.b64encode(image).decode()}]})