app.py

import gradio as gr
import os
import tempfile
from openai import OpenAI

# Set an environment variable for key
os.environ['OPENAI_API_KEY'] = os.environ.get('OPENAI_API_KEY')

client = OpenAI() # add api_key

def tts(text):
    response = client.audio.speech.create(
        model="tts-1",
        voice="alloy",
        input=text,
    )

    # Create a temp file to save the audio
    with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_file:
        temp_file.write(response.content)

    # Get the file path of the temp file
    temp_file_path = temp_file.name

    return temp_file_path


with gr.Blocks() as demo:
    gr.Markdown("# <center> OpenAI Text-To-Speech API with Gradio </center>")
    with gr.Row():
      dd1 = gr.Dropdown(choices=['tts-1','tts-1-hd'], label='Model')
      dd2 = gr.Dropdown(choices=['alloy', 'echo', 'fable', 'onyx', 'nova', 'shimmer'], label='Voice Options')

    text = gr.Textbox(label="Input text")
    btn = gr.Button("Text-To-Speech")
    output_audio = gr.Audio(label="Speech Output")
    
    btn.click(fn=tts, inputs=text, outputs=output_audio, api_name="tts", concurrency_limit=None)

demo.launch()