app.py

import gradio as gr
from huggingface_hub import InferenceClient
import os

listen_client = InferenceClient("openai/whisper-medium")

def audio_change(audio):
    return listen_client.automatic_speech_recognition(audio).text
    
gr.Interface(fn=audio_change, inputs=gr.Audio(label="Ses"), outputs=gr.Text()).launch()