实时音视频翻译-通义千问

如何使用

qwen3-livetranslate-flash-realtime 模型通过 WebSocket 协议接入，连接时需要以下配置项：

可通过以下 Python 示例代码建立连接。

# pip install websocket-client
import json
import websocket
import os

API_KEY=os.getenv("DASHSCOPE_API_KEY")
API_URL = "wss://dashscope.aliyuncs.com/api-ws/v1/realtime?model=qwen3-livetranslate-flash-realtime"

headers = [
    "Authorization: Bearer " + API_KEY
]

def on_open(ws):
    print(f"Connected to server: {API_URL}")
def on_message(ws, message):
    data = json.loads(message)
    print("Received event:", json.dumps(data, indent=2))
def on_error(ws, error):
    print("Error:", error)

ws = websocket.WebSocketApp(
    API_URL,
    header=headers,
    on_open=on_open,
    on_message=on_message,
    on_error=on_error
)

ws.run_forever()

支持的模型

qwen3-livetranslate-flash-realtime 是一款多语言音视频实时翻译模型，可识别 18 种语言，并实时翻译为 10 种语言的音频。

核心特性：

• 多语言支持：支持 18 种语言及 6 种汉语方言。包括中文、英语、法语、德语、俄语、日语、韩语等。支持普通话、粤语、四川话等方言。

• 视觉增强：利用视觉内容提升翻译准确性。模型通过分析口型、动作和画面中的文字，改善在嘈杂环境下或一词多义场景中的翻译效果。

• 3秒延迟：实现低至 3 秒的同传延迟。

• 无损同传：通过语义单元预测技术，解决跨语言语序问题。实时翻译质量接近离线翻译结果。

• 音色自然：生成音色自然的拟人语音。模型能根据源语音内容，自适应调节语气和情感。

快速开始

1. 准备运行环境

   您的 Python 版本需要不低于 3.10。

   首先安装 pyaudio。

   macOS

   brew install portaudio && pip install pyaudio

   Debian/Ubuntu

   sudo apt-get install python3-pyaudio
   
   或者
   
   pip install pyaudio

   CentOS

   sudo yum install -y portaudio portaudio-devel && pip install pyaudio

   Windows

   pip install pyaudio

   安装完成后，通过 pip 安装 websocket 相关的依赖：

   pip install websocket-client==1.8.0 websockets

2. 创建客户端

   在本地新建一个 python 文件，命名为livetranslate_client.py，并将以下代码复制进文件中：

   客户端代码-livetranslate_client.py

   import os
   import time
   import base64
   import asyncio
   import json
   import websockets
   import pyaudio
   import queue
   import threading
   import traceback
   
   class LiveTranslateClient:
       def __init__(self, api_key: str, target_language: str = "en", voice: str | None = "Cherry", *, audio_enabled: bool = True):
           if not api_key:
               raise ValueError("API key cannot be empty.")
               
           self.api_key = api_key
           self.target_language = target_language
           self.audio_enabled = audio_enabled
           self.voice = voice if audio_enabled else "Cherry"
           self.ws = None
           self.api_url = "wss://dashscope.aliyuncs.com/api-ws/v1/realtime?model=qwen3-livetranslate-flash-realtime"
           
           # 音频输入配置 (来自麦克风)
           self.input_rate = 16000
           self.input_chunk = 1600
           self.input_format = pyaudio.paInt16
           self.input_channels = 1
           
           # 音频输出配置 (用于播放)
           self.output_rate = 24000
           self.output_chunk = 2400
           self.output_format = pyaudio.paInt16
           self.output_channels = 1
           
           # 状态管理
           self.is_connected = False
           self.audio_player_thread = None
           self.audio_playback_queue = queue.Queue()
           self.pyaudio_instance = pyaudio.PyAudio()
   
       async def connect(self):
           """建立到翻译服务的 WebSocket 连接。"""
           headers = {"Authorization": f"Bearer {self.api_key}"}
           try:
               self.ws = await websockets.connect(self.api_url, additional_headers=headers)
               self.is_connected = True
               print(f"成功连接到服务器: {self.api_url}")
               await self.configure_session()
           except Exception as e:
               print(f"连接失败: {e}")
               self.is_connected = False
               raise
   
       async def configure_session(self):
           """配置翻译会话，设置目标语言、声音等。"""
           config = {
               "event_id": f"event_{int(time.time() * 1000)}",
               "type": "session.update",
               "session": {
                   # 'modalities' 控制输出类型。
                   # ["text", "audio"]: 同时返回翻译文本和合成音频（推荐）。
                   # ["text"]: 仅返回翻译文本。
                   "modalities": ["text", "audio"] if self.audio_enabled else ["text"],
                   **({"voice": self.voice} if self.audio_enabled and self.voice else {}),
                   "input_audio_format": "pcm16",
                   "output_audio_format": "pcm16",
                   "translation": {
                       "language": self.target_language
                   }
               }
           }
           print(f"发送会话配置: {json.dumps(config, indent=2, ensure_ascii=False)}")
           await self.ws.send(json.dumps(config))
   
       async def send_audio_chunk(self, audio_data: bytes):
           """将音频数据块编码并发送到服务器。"""
           if not self.is_connected:
               return
               
           event = {
               "event_id": f"event_{int(time.time() * 1000)}",
               "type": "input_audio_buffer.append",
               "audio": base64.b64encode(audio_data).decode()
           }
           await self.ws.send(json.dumps(event))
   
       def _audio_player_task(self):
           stream = self.pyaudio_instance.open(
               format=self.output_format,
               channels=self.output_channels,
               rate=self.output_rate,
               output=True,
               frames_per_buffer=self.output_chunk,
           )
           try:
               while self.is_connected or not self.audio_playback_queue.empty():
                   try:
                       audio_chunk = self.audio_playback_queue.get(timeout=0.1)
                       if audio_chunk is None: # 结束信号
                           break
                       stream.write(audio_chunk)
                       self.audio_playback_queue.task_done()
                   except queue.Empty:
                       continue
           finally:
               stream.stop_stream()
               stream.close()
   
       def start_audio_player(self):
           """启动音频播放线程（仅当启用音频输出时）。"""
           if not self.audio_enabled:
               return
           if self.audio_player_thread is None or not self.audio_player_thread.is_alive():
               self.audio_player_thread = threading.Thread(target=self._audio_player_task, daemon=True)
               self.audio_player_thread.start()
   
       async def handle_server_messages(self, on_text_received):
           """循环处理来自服务器的消息。"""
           try:
               async for message in self.ws:
                   event = json.loads(message)
                   event_type = event.get("type")
                   if event_type == "response.audio_transcript.delta":
                       text = event.get("transcript", "")
                       if text and on_text_received:
                           on_text_received(text)
                   
                   elif event_type == "response.audio.delta" and self.audio_enabled:
                       audio_b64 = event.get("delta", "")
                       if audio_b64:
                           audio_data = base64.b64decode(audio_b64)
                           self.audio_playback_queue.put(audio_data)
   
                   elif event_type == "response.done":
                       print("\n[INFO] 一轮响应完成。")
                       usage = event.get("response", {}).get("usage", {})
                       if usage:
                           print(f"[INFO] Token 使用情况: {json.dumps(usage, indent=2, ensure_ascii=False)}")
                   elif event_type == "response.audio_transcript.done":
                       print("\n[INFO] 翻译文本完成。")
                       text = event.get("transcript", "")
                       if text:
                           print(f"[INFO] 翻译文本: {text}")
                   elif event_type == "response.text.done":
                       print("\n[INFO] 翻译文本完成。")
                       text = event.get("text", "")
                       if text:
                           print(f"[INFO] 翻译文本: {text}")
   
           except websockets.exceptions.ConnectionClosed as e:
               print(f"[WARNING] 连接已关闭: {e}")
               self.is_connected = False
           except Exception as e:
               print(f"[ERROR] 消息处理时发生未知错误: {e}")
               traceback.print_exc()
               self.is_connected = False
   
       async def start_microphone_streaming(self):
           """从麦克风捕获音频并流式传输到服务器。"""
           stream = self.pyaudio_instance.open(
               format=self.input_format,
               channels=self.input_channels,
               rate=self.input_rate,
               input=True,
               frames_per_buffer=self.input_chunk
           )
           print("麦克风已启动，请开始说话...")
           try:
               while self.is_connected:
                   audio_chunk = await asyncio.get_event_loop().run_in_executor(
                       None, stream.read, self.input_chunk
                   )
                   await self.send_audio_chunk(audio_chunk)
           finally:
               stream.stop_stream()
               stream.close()
   
       async def close(self):
           """优雅地关闭连接和资源。"""
           self.is_connected = False
           if self.ws:
               await self.ws.close()
               print("WebSocket 连接已关闭。")
           
           if self.audio_player_thread:
               self.audio_playback_queue.put(None) # 发送结束信号
               self.audio_player_thread.join(timeout=1)
               print("音频播放线程已停止。")
               
           self.pyaudio_instance.terminate()
           print("PyAudio 实例已释放。")

3. 与模型互动

   在livetranslate_client.py的同级目录下新建另一个 python 文件，命名为main.py，并将以下代码复制进文件中：

   main.py

   import os
   import asyncio
   from livetranslate_client import LiveTranslateClient
   
   def print_banner():
       print("=" * 60)
       print("  基于通义千问 qwen3-livetranslate-flash-realtime")
       print("=" * 60 + "\n")
   
   def get_user_config():
       """获取用户配置"""
       print("请选择模式:")
       print("1. 语音+文本 [默认] | 2. 仅文本")
       mode_choice = input("请输入选项 (直接回车选择语音+文本): ").strip()
       audio_enabled = (mode_choice != "2")
   
       if audio_enabled:
           lang_map = {
               "1": "en", "2": "zh", "3": "ru", "4": "fr", "5": "de", "6": "pt",
               "7": "es", "8": "it", "9": "ko", "10": "ja", "11": "yue"
           }
           print("请选择翻译目标语言 (音频+文本 模式):")
           print("1. 英语 | 2. 中文 | 3. 俄语 | 4. 法语 | 5. 德语 | 6. 葡萄牙语 | 7. 西班牙语 | 8. 意大利语 | 9. 韩语 | 10. 日语 | 11. 粤语")
       else:
           lang_map = {
               "1": "en", "2": "zh", "3": "ru", "4": "fr", "5": "de", "6": "pt", "7": "es", "8": "it",
               "9": "id", "10": "ko", "11": "ja", "12": "vi", "13": "th", "14": "ar",
               "15": "yue", "16": "hi", "17": "el", "18": "tr"
           }
           print("请选择翻译目标语言 (仅文本 模式):")
           print("1. 英语 | 2. 中文 | 3. 俄语 | 4. 法语 | 5. 德语 | 6. 葡萄牙语 | 7. 西班牙语 | 8. 意大利语 | 9. 印尼语 | 10. 韩语 | 11. 日语 | 12. 越南语 | 13. 泰语 | 14. 阿拉伯语 | 15. 粤语 | 16. 印地语 | 17. 希腊语 | 18. 土耳其语")
   
       choice = input("请输入选项 (默认取第一个): ").strip()
       target_language = lang_map.get(choice, next(iter(lang_map.values())))
   
       voice = None
       if audio_enabled:
           print("\n请选择语音合成声音:")
           voice_map = {"1": "Cherry", "2": "Nofish", "3": "Sunny", "4": "Jada", "5": "Dylan", "6": "Peter", "7": "Eric", "8": "Kiki"}
           print("1. Cherry (女声) [默认] | 2. Nofish (男声) | 3. 晴儿 Sunny (四川女声) | 4. 阿珍 Jada (上海女声) | 5. 晓东 Dylan (北京男声) | 6. 李彼得 Peter (天津男声) | 7. 程川 Eric (四川男声) | 8. 阿清 Kiki (粤语女声)")
           voice_choice = input("请输入选项 (直接回车选择Cherry): ").strip()
           voice = voice_map.get(voice_choice, "Cherry")
       return target_language, voice, audio_enabled
   
   async def main():
       """主程序入口"""
       print_banner()
       
       api_key = os.environ.get("DASHSCOPE_API_KEY")
       if not api_key:
           print("[ERROR] 请设置环境变量 DASHSCOPE_API_KEY")
           print("  例如: export DASHSCOPE_API_KEY='your_api_key_here'")
           return
           
       target_language, voice, audio_enabled = get_user_config()
       print("\n配置完成:")
       print(f"  - 目标语言: {target_language}")
       if audio_enabled:
           print(f"  - 合成声音: {voice}")
       else:
           print("  - 输出模式: 仅文本")
       
       client = LiveTranslateClient(api_key=api_key, target_language=target_language, voice=voice, audio_enabled=audio_enabled)
       
       # 定义回调函数
       def on_translation_text(text):
           print(text, end="", flush=True)
   
       try:
           print("正在连接到翻译服务...")
           await client.connect()
           
           # 根据模式启动音频播放
           client.start_audio_player()
           
           print("\n" + "-" * 60)
           print("连接成功！请对着麦克风说话。")
           print("程序将实时翻译您的语音并播放结果。按 Ctrl+C 退出。")
           print("-" * 60 + "\n")
   
           # 并发运行消息处理和麦克风录音
           message_handler = asyncio.create_task(client.handle_server_messages(on_translation_text))
           tasks = [message_handler]
           # 无论是否启用音频输出，都需要从麦克风捕获音频进行翻译
           microphone_streamer = asyncio.create_task(client.start_microphone_streaming())
           tasks.append(microphone_streamer)
   
           await asyncio.gather(*tasks)
   
       except KeyboardInterrupt:
           print("\n\n用户中断，正在退出...")
       except Exception as e:
           print(f"\n发生严重错误: {e}")
       finally:
           print("\n正在清理资源...")
           await client.close()
           print("程序已退出。")
   
   if __name__ == "__main__":
       asyncio.run(main())

   运行main.py，通过麦克风说出要翻译的句子，模型会实时返回翻译完成的音频与文本。系统会检测您的音频起始位置并自动发送到服务器，无需手动发送。

视觉辅助翻译

qwen3-livetranslate-flash-realtime 模型可以接收图像输入，辅助音频翻译，适用于同音异义、低频专有名词识别场景。

将以下示例图片下载到本地：口罩.png面具.png

将以下代码下载到livetranslate_client.py同级目录并运行，向麦克风说"What is mask?"，在输入口罩图片时，模型会翻译为“什么是口罩？”；输入面具图片时，模型会翻译为“什么是面具？”。

import os
import time
import json
import asyncio
import contextlib
import functools

from livetranslate_client import LiveTranslateClient

# 可改为面具.png
IMAGE_PATH = "口罩.png"
RECORD_SECONDS = 6

def print_banner():
    print("=" * 60)
    print("  基于通义千问 qwen3-livetranslate-flash-realtime —— 单轮交互示例 (mask)")
    print("=" * 60 + "\n")

async def stream_microphone_once(client: LiveTranslateClient, image_bytes: bytes, duration: int = RECORD_SECONDS):
    pa = client.pyaudio_instance
    stream = pa.open(
        format=client.input_format,
        channels=client.input_channels,
        rate=client.input_rate,
        input=True,
        frames_per_buffer=client.input_chunk,
    )
    print(f"[INFO] 开始录音 {duration} 秒，请讲话……")
    start_t = time.time()
    loop = asyncio.get_event_loop()
    last_img_time = 0.0
    frame_interval = 0.5  # 2 fps
    try:
        while time.time() - start_t < duration and client.is_connected:
            data = await loop.run_in_executor(None, stream.read, client.input_chunk)
            await client.send_audio_chunk(data)

            # 每 0.5 秒追加一帧图片
            now = time.time()
            if now - last_img_time >= frame_interval:
                await client.send_image_frame(image_bytes)
                last_img_time = now
    finally:
        stream.stop_stream()
        stream.close()

    # 提交视频缓冲区（根据官方规范，仍使用 input_audio_buffer.commit）
    await client.ws.send(
        json.dumps({
            "event_id": f"event_{int(time.time()*1000)}",
            "type": "input_audio_buffer.commit",
        })
    )
    print("[INFO] 音频与图像缓冲区已提交。")

async def send_single_image(client: LiveTranslateClient, image_path: str):
    with open(image_path, "rb") as f:
        img_bytes = f.read()
    await client.send_image_frame(img_bytes)
    await client.ws.send(
        json.dumps({
            "event_id": f"event_{int(time.time()*1000)}",
            "type": "input_image_buffer.commit",
        })
    )
    print(f"[INFO] 已发送图像: {os.path.basename(image_path)}")

async def main():
    print_banner()
    api_key = os.environ.get("DASHSCOPE_API_KEY")
    if not api_key:
        print("[ERROR] 请先在环境变量 DASHSCOPE_API_KEY 中配置 API KEY")
        return

    client = LiveTranslateClient(api_key=api_key, target_language="zh", voice="Cherry", audio_enabled=True)

    def on_text(text: str):
        print(text, end="", flush=True)

    try:
        await client.connect()
        client.start_audio_player()
        message_task = asyncio.create_task(client.handle_server_messages(on_text))
        with open(IMAGE_PATH, "rb") as f:
            img_bytes = f.read()
        # 录音循环已包含图像帧，因此无需额外 send_single_image
        await stream_microphone_once(client, img_bytes, RECORD_SECONDS)
        await asyncio.sleep(15)
    finally:
        await client.close()
        if not message_task.done():
            message_task.cancel()
            with contextlib.suppress(asyncio.CancelledError):
                await message_task

if __name__ == "__main__":
    asyncio.run(main())

交互流程

实时语音翻译的交互流程遵循标准的 WebSocket 事件驱动模型，服务端自动检测语音起止并进行响应。

API 参考

请参见实时音视频翻译-通义千问。

计费说明

请参见模型列表与价格。

支持的语种

支持的音色