diff --git a/asr/asr_connector.py b/asr/asr_connector.py
index 1c52960..93d6e00 100644
--- a/asr/asr_connector.py
+++ b/asr/asr_connector.py
@@ -1,42 +1,53 @@
 #!/usr/bin/env python3
 
+import threading
+
 import rclpy
 from rclpy.node import Node
 from std_msgs.msg import String
 from rclpy.action import ActionClient
 from custom_interfaces.action import Prompt
 
+from TTS.api import TTS
+import sounddevice as sd
+
+
 class HighLevelPromptClient(Node):
     def __init__(self):
         super().__init__('high_level_prompt_client')
-        
-        # Subscriber to /high_level_prompt (std_msgs/String)
+
+        # Subscriber to /high_level_prompt
         self.subscription = self.create_subscription(
             String,
             '/high_level_prompt',
             self.prompt_callback,
             10
         )
-        self.subscription  # prevent unused variable warning
 
-        # Action client for /prompt_high_level (custom_interfaces/Prompt)
+        # Action client for /prompt_high_level
         self.action_client = ActionClient(self, Prompt, '/prompt_high_level')
 
+        # Load TTS once at startup
+        self.get_logger().info('Loading Coqui TTS model...')
+        self.tts = TTS("tts_models/en/ljspeech/tacotron2-DDC")
+        self.sample_rate = self.tts.synthesizer.output_sample_rate
+        self.get_logger().info(f'TTS loaded. sample_rate={self.sample_rate}')
+
+        # Prevent overlapping speech
+        self._speak_lock = threading.Lock()
+
     def prompt_callback(self, msg: String):
         self.get_logger().info(f'Received prompt: "{msg.data}"')
         self.send_prompt_action(msg.data)
 
     def send_prompt_action(self, prompt_text: str):
-        # Wait until the action server is available
         if not self.action_client.wait_for_server(timeout_sec=5.0):
             self.get_logger().error('Action server /prompt_high_level not available!')
             return
-        
-        # Create goal
+
         goal_msg = Prompt.Goal()
         goal_msg.prompt = prompt_text
-        
-        # Send goal asynchronously
+
         self._send_goal_future = self.action_client.send_goal_async(
             goal_msg,
             feedback_callback=self.feedback_callback
@@ -55,22 +66,49 @@ def goal_response_callback(self, future):
 
     def feedback_callback(self, feedback_msg):
         feedback = feedback_msg.feedback
-        self.get_logger().info(f'Feedback received: tools_called={feedback.tools_called}')
+        self.get_logger().info(
+            f'Feedback received: tools_called={feedback.tools_called}'
+        )
 
     def get_result_callback(self, future):
         result = future.result().result
-        self.get_logger().info(f'Action finished. Success: {result.success}, Final Response: "{result.final_response}"')
+        final_response = result.final_response.strip()
+
+        self.get_logger().info(
+            f'Action finished. Success: {result.success}, Final Response: "{final_response}"'
+        )
+
+        if result.success and final_response:
+            self.speak_text(final_response)
+
+    def speak_text(self, text: str):
+        # Run TTS in a background thread so ROS callbacks stay responsive
+        threading.Thread(
+            target=self._speak_text_worker,
+            args=(text,),
+            daemon=True
+        ).start()
+
+    def _speak_text_worker(self, text: str):
+        with self._speak_lock:
+            try:
+                self.get_logger().info(f'Speaking: "{text}"')
+                wav = self.tts.tts(text=text)
+                sd.stop()
+                sd.play(wav, samplerate=self.sample_rate)
+                sd.wait()
+            except Exception as e:
+                self.get_logger().error(f'TTS playback failed: {e}')
+
 
 def main(args=None):
     rclpy.init(args=args)
     node = HighLevelPromptClient()
+
     try:
         rclpy.spin(node)
     except KeyboardInterrupt:
         pass
     finally:
         node.destroy_node()
-        rclpy.shutdown()
-
-if __name__ == '__main__':
-    main()
+        rclpy.shutdown()
\ No newline at end of file
diff --git a/setup.py b/setup.py
index f636474..9cdbc98 100644
--- a/setup.py
+++ b/setup.py
@@ -17,7 +17,12 @@
         (os.path.join('share', package_name, 'launch'), 
             glob('launch/*.launch.py') + glob('launch/*.sh')),
     ],
-    install_requires=['setuptools'],
+    install_requires=[
+    'setuptools',
+    'coqui-tts[codec]',
+    'transformers==5.0.0',
+    'sounddevice',
+],
     zip_safe=True,
     maintainer='final-project',
     maintainer_email='karamahati@gmail.com',