app.py

import random
import gradio as gr
import numpy as np
from elevenlabs import voices, generate, set_api_key, UnauthenticatedRateLimitError

def pad_buffer(audio):
    buffer_size = len(audio)
    element_size = np.dtype(np.int16).itemsize
    if buffer_size % element_size != 0:
        audio = audio + b'\0' * (element_size - (buffer_size % element_size))
    return audio

def generate_voice(text, voice_name):
    try:
        audio = generate(
            text[:250],
            voice=voice_name,
            model="eleven_multilingual_v2"
        )
        return (44100, np.frombuffer(pad_buffer(audio), dtype=np.int16))
    except UnauthenticatedRateLimitError as e:
        raise gr.Error("Thanks for trying out ElevenLabs TTS! You've reached the free tier limit. Please provide an API key to continue.")
    except Exception as e:
        raise gr.Error(e)

badges = """
<div style="display: flex">
<!-- Badges hier -->
</div>
"""

description = """
<!-- Beschreibung hier -->
"""

with gr.Blocks(css="style.css", theme="NoCrypt/[email protected]") as block:
    gr.Markdown(badges)
    gr.Markdown(description)
    
    input_text = gr.Textbox(
        label="Input Text (250 characters max)",
        lines=2,
        value="Hello! 你好! Hola! नमस्ते! Bonjour! こんにちは! مرحبا! 안녕하세요! Ciao! Cześć! Привіт! Γειά σας! Здравей! வணக்கம்!",
        elem_id="input_text"
    )

    all_voices = voices()
    input_voice = gr.Dropdown(
        [voice.name for voice in all_voices],
        value="Bella",
        label="Voice",
        elem_id="input_voice"
    )

    run_button = gr.Button()

    out_audio = gr.Audio(
        label="Generated Voice",
        type="numpy",
        elem_id="out_audio",
        format="mp3"
    )

    inputs = [input_text, input_voice]
    outputs = [out_audio]

    run_button.click(
        fn=generate_voice,
        inputs=inputs,
        outputs=outputs,
        queue=True
    )

block.queue(concurrency_count=5).launch(debug=True)