nanmai-tts 2025.12.16.124832

Async Python library for Nanmai TTS (DeepSeek/Kimi voices)

nanmai-tts

Async Python library for Nanmai TTS (DeepSeek/Kimi voices).

This package provides an easy-to-use interface for the Nanmai AI text-to-speech service, featuring high-quality Chinese voices. It follows the same design pattern as edge-tts for consistency.

Installation

pip install nanmai-tts

Usage

Basic Usage

import asyncio
import nanmai_tts

async def main():
    # Create a Communicate instance
    communicate = nanmai_tts.Communicate("你好世界", "DeepSeek")

    # Get audio data
    audio_data = await communicate.get_audio_data()

    # Or save directly to file
    await communicate.save("output.mp3")

asyncio.run(main())

Streaming

async def stream_audio():
    communicate = nanmai_tts.Communicate("Hello world", "Kimi")

    async for chunk in communicate.stream():
        if chunk["type"] == "audio":
            # Process audio chunk
            print(f"Received {len(chunk['data'])} bytes of audio")

CLI Usage

nanmai-tts --text "你好世界" --voice DeepSeek --write-media output.mp3

Available Voices

• DeepSeek: Male voice, optimized for Chinese text
• Kimi: Female voice, optimized for Chinese text

API Reference

Communicate Class

• __init__(text: str, voice: str = "DeepSeek"): Initialize with text and voice
• stream(): Async generator yielding audio chunks
• save(filename: str): Save audio to file
• get_audio_data() -> bytes: Get complete audio data

Requirements

• Python 3.8+
• aiohttp

License

This project is for educational and research purposes. The underlying Nanmai TTS API is proprietary.

Disclaimer

This package uses reverse-engineered authentication methods for the Nanmai TTS service. Use at your own risk and in accordance with applicable laws and terms of service.