fix(azure): price azure_ai transcriptions at the azure_ai cost-map entry

litellm/llms/azure/audio_transcriptions.py

@@ -37,6 +37,7 @@ class AzureAudioTranscription(AzureChatCompletion):
         azure_ad_token: str | None = None,
         atranscription: bool = False,
         litellm_params: dict | None = None,
+        custom_llm_provider: str = "azure",
     ) -> TranscriptionResponse | Coroutine[Any, Any, TranscriptionResponse]:
         data: Final = {"model": model, "file": audio_file, **optional_params}
 
@@ -53,6 +54,7 @@ class AzureAudioTranscription(AzureChatCompletion):
                 logging_obj=logging_obj,
                 model=model,
                 litellm_params=litellm_params,
+                custom_llm_provider=custom_llm_provider,
             )
 
         azure_client: Final = self.get_azure_openai_client(
@@ -99,7 +101,7 @@ class AzureAudioTranscription(AzureChatCompletion):
             additional_args={"complete_input_dict": data},
             original_response=stringified_response,
         )
-        hidden_params: Final = {"model": model, "custom_llm_provider": "azure"}
+        hidden_params: Final = {"model": model, "custom_llm_provider": custom_llm_provider}
         final_response: Final[TranscriptionResponse] = convert_to_model_response_object(
             response_object=stringified_response,
             model_response_object=model_response,
@@ -122,6 +124,7 @@ class AzureAudioTranscription(AzureChatCompletion):
         client=None,
         max_retries=None,
         litellm_params: dict | None = None,
+        custom_llm_provider: str = "azure",
     ) -> TranscriptionResponse:
         response = None
         try:
@@ -178,7 +181,7 @@ class AzureAudioTranscription(AzureChatCompletion):
                 },
                 original_response=stringified_response,
             )
-            hidden_params: Final = {"model": model, "custom_llm_provider": "azure"}
+            hidden_params: Final = {"model": model, "custom_llm_provider": custom_llm_provider}
             response = convert_to_model_response_object(
                 _response_headers=headers,
                 response_object=stringified_response,

litellm/main.py

@@ -7805,6 +7805,7 @@ def transcription(
             azure_ad_token=azure_ad_token,
             max_retries=max_retries,
             litellm_params=litellm_params_dict,
+            custom_llm_provider=custom_llm_provider,
         )
     elif custom_llm_provider == "openai" or (custom_llm_provider in litellm.openai_compatible_providers):
         api_base = (

tests/test_litellm/llms/azure/test_audio_transcriptions.py

@@ -0,0 +1,61 @@
+import json
+from pathlib import Path
+from typing import Final
+
+import httpx
+import pytest
+from openai import AzureOpenAI
+
+import litellm
+from litellm.cost_calculator import completion_cost
+from litellm.litellm_core_utils.audio_utils.utils import calculate_request_duration
+
+AUDIO_FILE: Final = Path(__file__).parents[3] / "gettysburg.wav"
+WHISPER_COST_PER_SECOND: Final = 0.0001
+
+
+def _transcription_client() -> AzureOpenAI:
+    def handler(request: httpx.Request) -> httpx.Response:
+        return httpx.Response(200, json={"text": "Four score and seven years ago"})
+
+    return AzureOpenAI(
+        api_key="test-key",
+        api_version="2024-06-01",
+        azure_endpoint="https://example.cognitiveservices.azure.com",
+        http_client=httpx.Client(transport=httpx.MockTransport(handler)),
+    )
+
+
+def test_azure_ai_transcription_is_priced_at_the_azure_ai_entry():
+    with AUDIO_FILE.open("rb") as audio:
+        response = litellm.transcription(
+            model="azure_ai/whisper",
+            file=audio,
+            api_base="https://example.cognitiveservices.azure.com",
+            api_key="test-key",
+            api_version="2024-06-01",
+            client=_transcription_client(),
+        )
+    with AUDIO_FILE.open("rb") as audio:
+        duration = calculate_request_duration(audio)
+
+    assert duration is not None and duration > 0
+    assert response._hidden_params["custom_llm_provider"] == "azure_ai"
+    assert completion_cost(completion_response=response, call_type="transcription") == pytest.approx(
+        WHISPER_COST_PER_SECOND * duration
+    )
+
+
+def test_azure_transcription_keeps_the_azure_provider():
+    with AUDIO_FILE.open("rb") as audio:
+        response = litellm.transcription(
+            model="azure/whisper-1",
+            file=audio,
+            api_base="https://example.openai.azure.com",
+            api_key="test-key",
+            api_version="2024-06-01",
+            client=_transcription_client(),
+        )
+
+    assert response._hidden_params["custom_llm_provider"] == "azure"
+    assert json.loads(response.model_dump_json())["text"] == "Four score and seven years ago"