diff --git a/transcribebot.py b/transcribebot.py
index 303fa5e..e5a5222 100644
--- a/transcribebot.py
+++ b/transcribebot.py
@@ -2,7 +2,11 @@ import disnake
 from disnake.ext import commands
 from dotenv import load_dotenv
 import os
+from os import path
+from pydub import AudioSegment
+import speech_recognition as sr
 
+st = sr.Recognizer()
 load_dotenv()
 bot = commands.Bot(command_prefix='.')
 
@@ -10,6 +14,13 @@ bot = commands.Bot(command_prefix='.')
 async def transcribe(inter: disnake.ApplicationCommandInteraction, message: disnake.Message):
     # Reversing the message and sending it back.
     await message.attachments[0].save("audio.mp3")
-    await inter.response.send_message("audio saved", file=disnake.File("audio.mp3"))
+    mp3file = AudioSegment.from_mp3("audio.mp3")
+    mp3file.export("audio.wav", format="wav")
+    convertemessage = sr.AudioFile("audio.wav")
+    with convertemessage as sounds:
+        transcribeaudo = st.record(sounds)
+    await inter.response.send_message(st.recognize_sphinx(transcribeaudo), file=disnake.File("audio.mp3"))
+    os.remove("audio.mp3")
+    os.remove("audio.wav")
 
 bot.run(os.getenv("TOKEN"))
\ No newline at end of file