diff --git a/litellm/llms/azure/audio_transcription/transformation.py b/litellm/llms/azure/audio_transcription/transformation.py
index cda4f238263..e478c8ebf35 100644
--- a/litellm/llms/azure/audio_transcription/transformation.py
+++ b/litellm/llms/azure/audio_transcription/transformation.py
@@ -133,6 +133,17 @@ class AzureSpeechAudioTranscriptionConfig(BaseAudioTranscriptionConfig):
         raw_response: httpx.Response,
     ) -> TranscriptionResponse:
         response_json = raw_response.json()
+        recognition_status = response_json.get("RecognitionStatus")
+        if recognition_status is not None and recognition_status != "Success":
+            raise AzureSpeechAudioTranscriptionException(
+                message=(
+                    "Azure AI Speech transcription failed with "
+                    f"RecognitionStatus={recognition_status}."
+                ),
+                status_code=raw_response.status_code,
+                headers=raw_response.headers,
+            )
+
         text = self._extract_text(response_json)
         response = TranscriptionResponse(text=text)
         response._hidden_params = response_json
diff --git a/tests/test_litellm/llms/azure/test_azure_speech_audio_transcription.py b/tests/test_litellm/llms/azure/test_azure_speech_audio_transcription.py
index e42b304c7b7..6ed6be6f34f 100644
--- a/tests/test_litellm/llms/azure/test_azure_speech_audio_transcription.py
+++ b/tests/test_litellm/llms/azure/test_azure_speech_audio_transcription.py
@@ -183,6 +183,24 @@ def test_azure_speech_audio_transcription_response_transform(payload, expected_t
     assert result._hidden_params == payload
 
 
+def test_azure_speech_audio_transcription_response_raises_on_failed_status():
+    config = AzureSpeechAudioTranscriptionConfig()
+    response = httpx.Response(
+        200,
+        json={
+            "RecognitionStatus": "NoMatch",
+            "Offset": 0,
+            "Duration": 0,
+        },
+    )
+
+    with pytest.raises(
+        AzureSpeechAudioTranscriptionException,
+        match="RecognitionStatus=NoMatch",
+    ):
+        config.transform_audio_transcription_response(response)
+
+
 def test_azure_speech_transcription_routes_through_provider_config(monkeypatch):
     expected = TranscriptionResponse(text="hello")
     audio_handler = MagicMock(return_value=expected)