从零搭建一个免费的文本转语音在线工具（基于 Edge TTS）

浏览器 (HTML/CSS/JS)
    │
    │  POST /api/tts
    ▼
Flask 服务端 (server.py)
    │
    │  WebSocket
    ▼
Microsoft Edge TTS 服务
    │
    │  MP3 音频流
    ▼
浏览器播放/下载

#!/usr/bin/env python3
from flask import Flask, request, jsonify, send_from_directory
import edge_tts
import asyncio
import io

app = Flask(__name__, static_folder="web")

@app.route("/api/tts", methods=["POST"])
def tts():
    data = request.get_json()
    text = data.get("text", "").strip()
    voice = data.get("voice", "zh-CN-XiaoxiaoNeural")
    rate = data.get("rate", "+0%")
    pitch = data.get("pitch", "+0Hz")
    
    if not text:
        return jsonify(error="文本内容不能为空"), 400
    
    try:
        audio_data = asyncio.run(synthesize(text, voice, rate, pitch))
        return audio_data, 200, {"Content-Type": "audio/mpeg"}
    except Exception as e:
        return jsonify(error=f"语音合成失败: {e}"), 500

async def synthesize(text, voice, rate, pitch):
    communicate = edge_tts.Communicate(text, voice, rate=rate, pitch=pitch)
    buffer = io.BytesIO()
    async for chunk in communicate.stream():
        if chunk["type"] == "audio":
            buffer.write(chunk["data"])
    return buffer.getvalue()

if __name__ == "__main__":
    app.run(host="0.0.0.0", port=8000)

// 调用后端 TTS API
async function callEdgeTTS(text, voice, rate, pitch) {
    const ratePercent = Math.round((rate - 1) * 100);
    const rateStr = (ratePercent >= 0 ? '+' : '') + ratePercent + '%';
    const pitchStr = (pitch >= 0 ? '+' : '') + pitch + 'Hz';

    const response = await fetch('/api/tts', {
        method: 'POST',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify({
            text: text,
            voice: voice,
            rate: rateStr,
            pitch: pitchStr
        })
    });

    if (!response.ok) {
        throw new Error('语音合成请求失败');
    }

    return await response.blob();
}

// 播放音频
function playAudioBlob(blob) {
    const url = URL.createObjectURL(blob);
    const audio = document.getElementById('audioPlayer');
    audio.src = url;
    audio.play();
}

// 下载 MP3
function downloadAudio(blob) {
    const url = URL.createObjectURL(blob);
    const a = document.createElement('a');
    a.href = url;
    a.download = 'tts_audio.mp3';
    a.click();
}

const VOICES = {
    'zh-CN': {
        label: '中文(普通话)',
        voices: [
            { name: 'zh-CN-XiaoxiaoNeural', label: '晓晓', gender: '女' },
            { name: 'zh-CN-YunxiNeural', label: '云希', gender: '男' },
            { name: 'zh-CN-XiaoyiNeural', label: '晓伊', gender: '女' },
            { name: 'zh-CN-YunjianNeural', label: '云健', gender: '男' },
        ]
    },
    'en-US': {
        label: '英文(美国)',
        voices: [
            { name: 'en-US-JennyNeural', label: 'Jenny', gender: '女' },
            { name: 'en-US-GuyNeural', label: 'Guy', gender: '男' },
        ]
    },
    // ... 更多语言
};

async function synthesize() {
    try {
        // 优先使用 Edge TTS
        const blob = await callEdgeTTS(text, voice, rate, pitch);
        playAudioBlob(blob);
        enableDownload(); // 支持下载
    } catch (edgeError) {
        // 备用方案：Web Speech API
        console.warn('Edge TTS 不可用，切换到浏览器内置语音');
        const utterance = new SpeechSynthesisUtterance(text);
        utterance.rate = rate;
        speechSynthesis.speak(utterance);
        disableDownload(); // 不支持下载
    }
}

# 限制单次文本长度
MAX_TEXT_LENGTH = 1000

# 限制每小时请求次数
MAX_REQUESTS_PER_HOUR = 10

# 按 IP 记录请求
request_log = {}

def check_rate_limit(ip):
    now = time.time()
    timestamps = request_log.get(ip, [])
    timestamps = [t for t in timestamps if t > now - 3600]
    request_log[ip] = timestamps
    return len(timestamps) < MAX_REQUESTS_PER_HOUR

# 安装依赖
pip install flask edge-tts

# 启动服务
python server.py

# 访问 http://localhost:8000

server {
    listen 80;
    server_name text2voice.cc;

    location / {
        proxy_pass http://127.0.0.1:8000;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
    }
}

nohup python3 server.py > server.log 2>&1 &

web_text2voice/
├── web/
│   ├── index.html      # 主页面
│   ├── about.html      # 关于页面
│   ├── privacy.html    # 隐私政策
│   ├── css/
│   │   └── style.css   # 样式文件
│   └── js/
│       └── app.js      # 前端逻辑
├── server.py           # Flask 后端服务
└── README.md           # 项目文档

audio_data = asyncio.run(synthesize(text, voice, rate, pitch))

async def synthesize(text, voice, rate, pitch):
    communicate = edge_tts.Communicate(text, voice, rate=rate, pitch=pitch)
    buffer = io.BytesIO()
    async for chunk in communicate.stream():
        if chunk["type"] == "audio":
            buffer.write(chunk["data"])
    return buffer.getvalue()

@app.after_request
def after_request(response):
    response.headers.add('Access-Control-Allow-Origin', '*')
    return response