app.py

import gradio as gr
from kokoro import KPipeline
import soundfile as sf
import os

# Initialize the pipeline
pipeline = KPipeline(lang_code='a')

def generate_speech(text, voice):
    # Ensure text is not empty
    if not text:
        return None
        
    # Generate speech
    generator = pipeline(text, voice=voice, speed=1, split_pattern=r'\n+')
    
    # Save the output
    output_path = "output.wav"
    for _, _, audio in generator:
        sf.write(output_path, audio, 24000)
        return output_path
    return None

# Create the Gradio interface
# The api_name parameter enables the API endpoint automatically
demo = gr.Interface(
    fn=generate_speech,
    inputs=[
        gr.Textbox(label="Enter Text", placeholder="Type your text here..."),
        gr.Dropdown(choices=["af_sarah", "am_michael", "af_bella"], label="Select Voice", value="af_sarah")
    ],
    outputs=gr.Audio(label="Generated Audio"),
    title="TTS Generator & API",
    description="Type text to generate audio. This app also functions as an API.",
    api_name="generate_audio"
)

if __name__ == "__main__":
    demo.launch()