test(fireworks): mock whisper transcription tests instead of live calls (#30391)

TestFireworksAIAudioTranscription inherited two tests from
BaseLLMAudioTranscriptionTest that made real calls to
audio-prod.api.fireworks.ai. The audio host is a separately-entitled
product, and the CI key gets a 401 there, so both tests fail without
signalling anything about litellm.

Override them in the subclass with a dependency-injected mock OpenAI
client (the openai-compatible path uses the openai SDK, which is why the
failure surfaced as openai.AuthenticationError). The real routing,
model-prefix stripping, and response parsing still run; only the network
call is mocked. The sync path mocks audio.transcriptions.create and the
async path mocks audio.transcriptions.with_raw_response.create.

This matches the existing direction in this file, where the
document-inlining and global-disable tests were already converted off
live Fireworks calls.

tests/llm_translation/test_fireworks_ai_translation.py

@@ -70,6 +70,11 @@ def test_map_response_format():
     assert result == {"response_format": response_format}
 
 
+_AUDIO_FILE_PATH = os.path.join(
+    os.path.dirname(os.path.realpath(__file__)), "gettysburg.wav"
+)
+
+
 class TestFireworksAIAudioTranscription(BaseLLMAudioTranscriptionTest):
     def get_base_audio_transcription_call_args(self) -> dict:
         return {
@@ -80,6 +85,60 @@ class TestFireworksAIAudioTranscription(BaseLLMAudioTranscriptionTest):
     def get_custom_llm_provider(self) -> litellm.LlmProviders:
         return litellm.LlmProviders.FIREWORKS_AI
 
+    def test_audio_transcription(self):
+        from unittest.mock import MagicMock
+
+        from openai.types.audio import Transcription
+
+        audio_file = open(_AUDIO_FILE_PATH, "rb")
+        mock_client = MagicMock()
+        mock_client.audio.transcriptions.create.return_value = Transcription(
+            text="four score and seven years ago"
+        )
+
+        transcript = transcription(
+            **self.get_base_audio_transcription_call_args(),
+            file=audio_file,
+            api_key="fw-test-key",
+            client=mock_client,
+        )
+
+        assert transcript.text == "four score and seven years ago"
+        sent = mock_client.audio.transcriptions.create.call_args.kwargs
+        assert sent["model"] == "whisper-v3"
+        assert sent["file"] is audio_file
+
+    @pytest.mark.asyncio
+    async def test_audio_transcription_async(self):
+        from unittest.mock import AsyncMock, MagicMock
+
+        from openai.types.audio import Transcription
+
+        audio_file = open(_AUDIO_FILE_PATH, "rb")
+        raw_response = MagicMock()
+        raw_response.headers = {}
+        raw_response.parse.return_value = Transcription(
+            text="four score and seven years ago"
+        )
+        mock_client = MagicMock()
+        mock_client.audio.transcriptions.with_raw_response.create = AsyncMock(
+            return_value=raw_response
+        )
+
+        transcript = await litellm.atranscription(
+            **self.get_base_audio_transcription_call_args(),
+            file=audio_file,
+            api_key="fw-test-key",
+            client=mock_client,
+        )
+
+        assert transcript.text == "four score and seven years ago"
+        sent = (
+            mock_client.audio.transcriptions.with_raw_response.create.call_args.kwargs
+        )
+        assert sent["model"] == "whisper-v3"
+        assert sent["file"] is audio_file
+
 
 @pytest.mark.parametrize(
     "disable_add_transform_inline_image_block",