feat: 接入豆包 Seed-ICL 2.0 语音合成引擎

- 新增 doubao_service.py 调用火山引擎 TTS API
- 路由加 doubao 引擎分支
- 前端引擎下拉菜单加豆包选项

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

backend/app/routes/tts_synthesize.py

@@ -3,6 +3,7 @@ import base64
 import os
 from app.services.minimax_service import synthesize_minimax
 from app.services.cosyvoice_service import synthesize_cosyvoice
+from app.services.doubao_service import synthesize_doubao
 from app.routes.auth import login_required
 from app.db import get_db
 
@@ -37,6 +38,11 @@ def synthesize():
                 text=text, speed=speed, pitch=pitch,
                 volume=volume, instruction=instruction
             )
+        elif engine == 'doubao':
+            audio_data = synthesize_doubao(
+                text=text, speed=speed, pitch=pitch,
+                volume=volume, emotion=emotion
+            )
         else:
             audio_data = synthesize_minimax(
                 text=text, speed=speed, pitch=pitch,

backend/app/services/doubao_service.py

@@ -0,0 +1,52 @@
+import os
+import uuid
+import requests
+
+def synthesize_doubao(text, speed=1.0, pitch=0, volume=1.0, emotion='neutral'):
+    """调用火山引擎豆包 Seed-ICL 2.0 语音合成 API"""
+    api_key = os.getenv('DOUBAO_API_KEY', '')
+    if not api_key:
+        raise ValueError('DOUBAO_API_KEY未设置')
+
+    url = 'https://openspeech.bytedance.com/api/v1/tts'
+    headers = {
+        'x-api-key': api_key,
+        'Content-Type': 'application/json'
+    }
+
+    payload = {
+        'app': {
+            'cluster': 'volcano_icl'
+        },
+        'user': {
+            'uid': 'military_tech_voice'
+        },
+        'audio': {
+            'voice_type': 'S_MK0j18H72',
+            'encoding': 'mp3',
+            'speed_ratio': speed,
+        },
+        'request': {
+            'reqid': str(uuid.uuid4()),
+            'text': text,
+            'operation': 'query'
+        }
+    }
+
+    response = requests.post(url, json=payload, headers=headers, timeout=60)
+
+    if response.status_code != 200:
+        raise Exception(f'豆包API错误: {response.status_code} - {response.text}')
+
+    result = response.json()
+
+    # 豆包返回格式: {"code": 3000, "data": "base64编码的音频"}
+    if result.get('code') == 3000 and result.get('data'):
+        import base64
+        audio_bytes = base64.b64decode(result['data'])
+        return audio_bytes
+
+    # 错误处理
+    error_msg = result.get('message', '未知错误')
+    error_code = result.get('code', 'unknown')
+    raise Exception(f'豆包返回错误 [{error_code}]: {error_msg}')

frontend/app.js

@@ -17,6 +17,11 @@ const ENGINE_CONFIG = {
         name: 'CosyVoice',
         emotionMode: 'global',
         color: '#ff6b6b'
+    },
+    doubao: {
+        name: '豆包',
+        emotionMode: 'global',
+        color: '#7c5cfc'
     }
 };
 

frontend/index.html

@@ -40,6 +40,7 @@
                         <select id="engine-select">
                             <option value="minimax">MiniMax</option>
                             <option value="cosyvoice">CosyVoice</option>
+                            <option value="doubao">豆包</option>
                         </select>
                         <span class="toolbar-divider"></span>
                         <span class="toolbar-label">情绪：</span>