Merge pull request #4024 from pipecat-ai/mb/fix-lang-enum-stt-tts

markbackman · web-flow · commit 53f49ac0943d · 2026-03-16T21:08:48.000-04:00
diff --git a/changelog/4024.fixed.md b/changelog/4024.fixed.md
@@ -0,0 +1 @@
+- Fixed `Language` enum values (e.g. `Language.ES`) not being converted to service-specific codes when passed via `settings=Service.Settings(language=Language.ES)` at init time. This caused API errors (e.g. 400 from Rime) because the raw enum was sent instead of the expected language code (e.g. `"spa"`). Runtime updates via `UpdateSettingsFrame` were unaffected. The fix centralizes conversion in the base `TTSService` and `STTService` classes so all services handle this consistently.
diff --git a/src/pipecat/services/asyncai/tts.py b/src/pipecat/services/asyncai/tts.py
@@ -171,9 +171,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    self.language_to_service_language(params.language) if params.language else None
-                )
+                default_settings.language = params.language
 
         # 4. Apply settings delta (canonical API, always wins)
         if settings is not None:
@@ -565,9 +563,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    self.language_to_service_language(params.language) if params.language else None
-                )
+                default_settings.language = params.language
 
         # 4. Apply settings delta (canonical API, always wins)
         if settings is not None:
diff --git a/src/pipecat/services/aws/stt.py b/src/pipecat/services/aws/stt.py
@@ -100,13 +100,13 @@ def __init__(
         # 1. Initialize default_settings with hardcoded defaults
         default_settings = self.Settings(
             model=None,
-            language=self.language_to_service_language(Language.EN),
+            language=Language.EN,
         )
 
         # 2. Apply direct init arg overrides (deprecated)
         if language is not None:
             self._warn_init_param_moved_to_settings("language", "language")
-            default_settings.language = self.language_to_service_language(language)
+            default_settings.language = language
 
         # 3. (No step 3, as there's no params object to apply)
 
diff --git a/src/pipecat/services/aws/tts.py b/src/pipecat/services/aws/tts.py
@@ -230,11 +230,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 default_settings.engine = params.engine
-                default_settings.language = (
-                    self.language_to_service_language(params.language)
-                    if params.language
-                    else "en-US"
-                )
+                default_settings.language = params.language if params.language else "en-US"
                 default_settings.pitch = params.pitch
                 default_settings.rate = params.rate
                 default_settings.volume = params.volume
diff --git a/src/pipecat/services/azure/stt.py b/src/pipecat/services/azure/stt.py
@@ -106,13 +106,13 @@ def __init__(
         # 1. Initialize default_settings with hardcoded defaults
         default_settings = self.Settings(
             model=None,
-            language=language_to_azure_language(Language.EN_US),
+            language=Language.EN_US,
         )
 
         # 2. Apply direct init arg overrides (deprecated)
         if language is not None and language != Language.EN_US:
             self._warn_init_param_moved_to_settings("language", "language")
-            default_settings.language = language_to_azure_language(language)
+            default_settings.language = language
 
         # 3. (No step 3, as there's no params object to apply)
 
diff --git a/src/pipecat/services/azure/tts.py b/src/pipecat/services/azure/tts.py
@@ -312,11 +312,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 default_settings.emphasis = params.emphasis
-                default_settings.language = (
-                    self.language_to_service_language(params.language)
-                    if params.language
-                    else "en-US"
-                )
+                default_settings.language = params.language if params.language else "en-US"
                 default_settings.pitch = params.pitch
                 default_settings.rate = params.rate
                 default_settings.role = params.role
@@ -809,11 +805,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 default_settings.emphasis = params.emphasis
-                default_settings.language = (
-                    self.language_to_service_language(params.language)
-                    if params.language
-                    else "en-US"
-                )
+                default_settings.language = params.language if params.language else "en-US"
                 default_settings.pitch = params.pitch
                 default_settings.rate = params.rate
                 default_settings.role = params.role
diff --git a/src/pipecat/services/camb/tts.py b/src/pipecat/services/camb/tts.py
@@ -260,9 +260,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = (
-                        self.language_to_service_language(params.language) or "en-us"
-                    )
+                    default_settings.language = params.language
                 if params.user_instructions is not None:
                     default_settings.user_instructions = params.user_instructions
 
diff --git a/src/pipecat/services/cartesia/tts.py b/src/pipecat/services/cartesia/tts.py
@@ -302,7 +302,7 @@ def __init__(
         default_settings = self.Settings(
             model="sonic-3",
             voice=None,
-            language=language_to_cartesia_language(Language.EN),
+            language=Language.EN,
             generation_config=None,
             pronunciation_dict_id=None,
         )
@@ -320,7 +320,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.generation_config is not None:
                     default_settings.generation_config = params.generation_config
                 if params.pronunciation_dict_id is not None:
@@ -749,7 +749,7 @@ def __init__(
         default_settings = self.Settings(
             model="sonic-3",
             voice=None,
-            language=language_to_cartesia_language(Language.EN),
+            language=Language.EN,
             generation_config=None,
             pronunciation_dict_id=None,
         )
@@ -767,7 +767,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.generation_config is not None:
                     default_settings.generation_config = params.generation_config
                 if params.pronunciation_dict_id is not None:
diff --git a/src/pipecat/services/elevenlabs/stt.py b/src/pipecat/services/elevenlabs/stt.py
@@ -272,7 +272,7 @@ def __init__(
         # 1. Initialize default_settings with hardcoded defaults
         default_settings = self.Settings(
             model="scribe_v2",
-            language=language_to_elevenlabs_language(Language.EN),
+            language=Language.EN,
             tag_audio_events=None,
         )
 
@@ -286,7 +286,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = language_to_elevenlabs_language(params.language)
+                    default_settings.language = params.language
                 default_settings.tag_audio_events = params.tag_audio_events
 
         # 4. Apply settings delta (canonical API, always wins)
diff --git a/src/pipecat/services/elevenlabs/tts.py b/src/pipecat/services/elevenlabs/tts.py
@@ -449,7 +449,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.stability is not None:
                     default_settings.stability = params.stability
                 if params.similarity_boost is not None:
@@ -1014,7 +1014,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.optimize_streaming_latency is not None:
                     default_settings.optimize_streaming_latency = params.optimize_streaming_latency
                 if params.stability is not None:
diff --git a/src/pipecat/services/fal/stt.py b/src/pipecat/services/fal/stt.py
@@ -216,7 +216,7 @@ def __init__(
         # 1. Initialize default_settings with hardcoded defaults
         default_settings = self.Settings(
             model=None,
-            language=language_to_fal_language(Language.EN),
+            language=Language.EN,
         )
 
         # 2. (no deprecated direct args for this service)
@@ -226,7 +226,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = language_to_fal_language(params.language)
+                    default_settings.language = params.language
                 if params.task != "transcribe":
                     task = params.task
                 if params.chunk_level != "segment":
diff --git a/src/pipecat/services/google/tts.py b/src/pipecat/services/google/tts.py
@@ -653,7 +653,7 @@ def __init__(
                 if params.emphasis is not None:
                     default_settings.emphasis = params.emphasis
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.gender is not None:
                     default_settings.gender = params.gender
                 if params.google_style is not None:
@@ -1090,7 +1090,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.speaking_rate is not None:
                     default_settings.speaking_rate = params.speaking_rate
 
@@ -1346,7 +1346,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.prompt is not None:
                     default_settings.prompt = params.prompt
                 if params.multi_speaker is not None:
diff --git a/src/pipecat/services/groq/stt.py b/src/pipecat/services/groq/stt.py
@@ -85,7 +85,7 @@ def __init__(
         # --- 1. Hardcoded defaults ---
         default_settings = self.Settings(
             model="whisper-large-v3-turbo",
-            language=self.language_to_service_language(Language.EN),
+            language=Language.EN,
             prompt=None,
             temperature=None,
         )
@@ -96,7 +96,7 @@ def __init__(
             default_settings.model = model
         if language is not None:
             self._warn_init_param_moved_to_settings("language", "language")
-            default_settings.language = self.language_to_service_language(language)
+            default_settings.language = language
         if prompt is not None:
             self._warn_init_param_moved_to_settings("prompt", "prompt")
             default_settings.prompt = prompt
diff --git a/src/pipecat/services/kokoro/tts.py b/src/pipecat/services/kokoro/tts.py
@@ -150,7 +150,7 @@ def __init__(
         default_settings = self.Settings(
             model=None,
             voice=None,
-            language=language_to_kokoro_language(Language.EN),
+            language=Language.EN,
         )
 
         # 2. Apply direct init arg overrides (deprecated)
@@ -162,7 +162,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = language_to_kokoro_language(params.language)
+                default_settings.language = params.language
 
         # 4. Apply settings delta (canonical API, always wins)
         if settings is not None:
diff --git a/src/pipecat/services/lmnt/tts.py b/src/pipecat/services/lmnt/tts.py
@@ -114,6 +114,10 @@ def __init__(
 
             sample_rate: Audio sample rate. If None, uses default.
             language: Language for synthesis. Defaults to English.
+
+                .. deprecated:: 0.0.106
+                    Use ``settings=LmntTTSService.Settings(language=...)`` instead.
+
             output_format: Audio output format. One of "pcm_s16le", "pcm_f32le",
                 "mp3", "ulaw", "webm". Defaults to "pcm_s16le".
             model: TTS model to use.
@@ -129,13 +133,16 @@ def __init__(
         default_settings = self.Settings(
             model="aurora",
             voice=None,
-            language=self.language_to_service_language(language),
+            language=Language.EN,
         )
 
         # 2. Apply direct init arg overrides (deprecated)
         if voice_id is not None:
             self._warn_init_param_moved_to_settings("voice_id", "voice")
             default_settings.voice = voice_id
+        if language is not None:
+            self._warn_init_param_moved_to_settings("language", "language")
+            default_settings.language = language
         if model is not None:
             self._warn_init_param_moved_to_settings("model", "model")
             default_settings.model = model
diff --git a/src/pipecat/services/neuphonic/tts.py b/src/pipecat/services/neuphonic/tts.py
@@ -153,7 +153,7 @@ def __init__(
         default_settings = self.Settings(
             model=None,
             voice=None,
-            language=self.language_to_service_language(Language.EN),
+            language=Language.EN,
             speed=1.0,
         )
 
@@ -167,7 +167,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.speed is not None:
                     default_settings.speed = params.speed
 
@@ -487,7 +487,7 @@ def __init__(
         default_settings = self.Settings(
             model=None,
             voice=None,
-            language=self.language_to_service_language(Language.EN),
+            language=Language.EN,
             speed=1.0,
         )
 
@@ -501,7 +501,7 @@ def __init__(
             self._warn_init_param_moved_to_settings("params")
             if not settings:
                 if params.language is not None:
-                    default_settings.language = self.language_to_service_language(params.language)
+                    default_settings.language = params.language
                 if params.speed is not None:
                     default_settings.speed = params.speed
 
diff --git a/src/pipecat/services/nvidia/stt.py b/src/pipecat/services/nvidia/stt.py
@@ -503,7 +503,7 @@ def __init__(
         # 1. Initialize default_settings with hardcoded defaults
         default_settings = self.Settings(
             model=model_function_map.get("model_name"),
-            language=language_to_nvidia_riva_language(Language.EN_US) or "en-US",
+            language=Language.EN_US,
             profanity_filter=False,
             automatic_punctuation=True,
             verbatim_transcripts=False,
@@ -517,9 +517,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    language_to_nvidia_riva_language(params.language or Language.EN_US) or "en-US"
-                )
+                default_settings.language = params.language or Language.EN_US
                 default_settings.profanity_filter = params.profanity_filter
                 default_settings.automatic_punctuation = params.automatic_punctuation
                 default_settings.verbatim_transcripts = params.verbatim_transcripts
diff --git a/src/pipecat/services/openai/stt.py b/src/pipecat/services/openai/stt.py
@@ -119,7 +119,7 @@ def __init__(
         _language = language or Language.EN
         default_settings = self.Settings(
             model="gpt-4o-transcribe",
-            language=self.language_to_service_language(_language),
+            language=_language,
             prompt=None,
             temperature=None,
         )
diff --git a/src/pipecat/services/rime/tts.py b/src/pipecat/services/rime/tts.py
@@ -251,9 +251,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    self.language_to_service_language(params.language) if params.language else None
-                )
+                default_settings.language = params.language
                 default_settings.segment = params.segment
                 default_settings.speedAlpha = params.speed_alpha
                 # Arcana params
@@ -754,9 +752,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    self.language_to_service_language(params.language) if params.language else "eng"
-                )
+                default_settings.language = params.language
                 default_settings.speedAlpha = params.speed_alpha
                 default_settings.reduceLatency = params.reduce_latency
                 default_settings.pauseBetweenBrackets = params.pause_between_brackets
@@ -984,9 +980,7 @@ def __init__(
         if params is not None:
             self._warn_init_param_moved_to_settings("params")
             if not settings:
-                default_settings.language = (
-                    self.language_to_service_language(params.language) if params.language else None
-                )
+                default_settings.language = params.language
                 default_settings.segment = params.segment
                 default_settings.repetition_penalty = params.repetition_penalty
                 default_settings.temperature = params.temperature
diff --git a/src/pipecat/services/sambanova/stt.py b/src/pipecat/services/sambanova/stt.py
diff --git a/src/pipecat/services/sarvam/tts.py b/src/pipecat/services/sarvam/tts.py
diff --git a/src/pipecat/services/stt_service.py b/src/pipecat/services/stt_service.py
diff --git a/src/pipecat/services/tts_service.py b/src/pipecat/services/tts_service.py
diff --git a/src/pipecat/services/whisper/base_stt.py b/src/pipecat/services/whisper/base_stt.py
diff --git a/src/pipecat/services/xtts/tts.py b/src/pipecat/services/xtts/tts.py

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	+- Fixed `Language` enum values (e.g. `Language.ES`) not being converted to service-specific codes when passed via `settings=Service.Settings(language=Language.ES)` at init time. This caused API errors (e.g. 400 from Rime) because the raw enum was sent instead of the expected language code (e.g. `"spa"`). Runtime updates via `UpdateSettingsFrame` were unaffected. The fix centralizes conversion in the base `TTSService` and `STTService` classes so all services handle this consistently.