diff --git a/LLM/language_model.py b/LLM/language_model.py
index e0d5829a138d87135690b8aca0d5a5db4093e7c3..cb24c895d04e912d9ca06092a082f29b9d01b2b1 100644
--- a/LLM/language_model.py
+++ b/LLM/language_model.py
@@ -101,8 +101,11 @@ class LanguageModelHandler(BaseHandler):
                 f"{self.__class__.__name__}:  warmed up! time: {start_event.elapsed_time(end_event) * 1e-3:.3f} s"
             )
 
-    def process(self, prompt, language_id=None):
+    def process(self, prompt):
         logger.debug("infering language model...")
+        language_id = None
+        if isinstance(prompt, tuple):
+            prompt, language_id = prompt
 
         self.chat.append({"role": self.user_role, "content": prompt})
         thread = Thread(
diff --git a/TTS/melo_handler.py b/TTS/melo_handler.py
index 6b5d83147e16e6c1fdf87c3d0f45daffc09f1aa0..45eef88f2297b946d6602961b13ec6ea1e3987c1 100644
--- a/TTS/melo_handler.py
+++ b/TTS/melo_handler.py
@@ -52,8 +52,11 @@ class MeloTTSHandler(BaseHandler):
         logger.info(f"Warming up {self.__class__.__name__}")
         _ = self.model.tts_to_file("text", self.speaker_id, quiet=True)
 
-    def process(self, llm_sentence, language_id=None):
+    def process(self, llm_sentence):
         console.print(f"[green]ASSISTANT: {llm_sentence}")
+        language_id = None
+        if isinstance(prompt, tuple):
+            prompt, language_id = prompt
 
         if language_id is not None and self.language != language_id:
             self.model = TTS(language=WHISPER_LANGUAGE_TO_MELO_LANGUAGE[self.language], device=self.device)