diff --git a/stt_voice_messages/stt_voice_messages.py b/stt_voice_messages/stt_voice_messages.py
index bbc8c68..ff82093 100644
--- a/stt_voice_messages/stt_voice_messages.py
+++ b/stt_voice_messages/stt_voice_messages.py
@@ -40,7 +40,7 @@ class STTVoiceMessagesPlugin(GajimPlugin):
 
         self.config_default_values = {
             'use_multilanguage_model': (True, ''),
-            'model_size': ('tiny', '')
+            'model_size': ('base', '')
         }
 
         self._audio_file = None
@@ -77,7 +77,11 @@ class STTBox(Gtk.Box):
         self._audio_file = audio_file
 
         self._transcribe_button = Gtk.Button(label=_('Transcribe'))
-        self._transcription_label = Gtk.Label(label='Fooo')
+
+        self._transcription_label = Gtk.Label(label=_('Nothing transcribed yet'))
+        self._transcription_label.set_max_width_chars(40)
+        self._transcription_label.set_line_wrap(True)
+
         self.add(self._transcribe_button)
         self.add(self._transcription_label)
 
@@ -88,7 +92,10 @@ class STTBox(Gtk.Box):
 
     def _on_transcribe_clicked(self, _button: Gtk.Button):
         text = self._trascribe_by_whisper()
-        self._transcription_label.set_text(text)
+        if text.strip() != '':
+            self._transcription_label.set_text(text.strip())
+        else:
+            self._transcription_label.set_text(_('_Have not heard any word!_'))
 
     def _trascribe_by_whisper(self) -> str:
         model = whisper.load_model(self._config['model_size'])