elevenlabs-dialogue

安装 belt CLI 技能： npx skills add belt-sh/cli

ElevenLabs 对话

通过 inference.sh CLI 生成多人对话音频。

快速开始

需要 inference.sh CLI（belt）。安装说明

belt login

# Generate dialogue
belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "Have you tried the new feature?", "voice": "george"},
    {"text": "Not yet, but I heard it is amazing.", "voice": "aria"},
    {"text": "You should check it out today.", "voice": "george"}
  ]
}'

语音选项

每位说话者都有 22 种以上的高级语音可选：

热门搭配

搭配	语音	适用场景
访谈	george + aria	专业问答
闲聊	brian + sarah	轻松对话
教程	daniel + jessica	教学讲解
辩论	adam + alice	不同观点
播客	charlie + bella	娱乐

所有语音

女性：aria、alice、bella、jessica、laura、lily、sarah、matilda

男性：george、adam、bill、brian、callum、charlie、chris、daniel、eric、harry、liam、river、roger、will

语音指导

在方括号中添加指导语，以控制表达方式：

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "[excitedly] Guess what happened today!", "voice": "aria"},
    {"text": "[curiously] What? Tell me!", "voice": "george"},
    {"text": "[proudly] We hit ten thousand users!", "voice": "aria"},
    {"text": "[surprised] No way, that is incredible!", "voice": "george"}
  ]
}'

指导关键词

指导语	效果
[excitedly]	充满活力、积极热情
[sadly]	忧郁、富有情感
[whispering]	轻柔、低声说话
[angrily]	强烈、有力地表达
[sarcastically]	带有讽刺意味的语调
[curiously]	疑问、好奇
[proudly]	自信、成就感十足
[nervously]	犹豫、不确定
[cheerfully]	快乐、明快

示例

播客节目

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "Welcome back to Tech Talk! Today we are discussing the latest in AI.", "voice": "george"},
    {"text": "Thanks for having me. This is such an exciting topic right now.", "voice": "aria"},
    {"text": "So let us start with the big question. How is AI changing creative work?", "voice": "george"},
    {"text": "Great question. I think the biggest shift is in accessibility. Tools that used to require specialized skills are now available to everyone.", "voice": "aria"},
    {"text": "Can you give us a specific example?", "voice": "george"},
    {"text": "Sure. Take audio production. A year ago, you needed a studio and voice actors. Now you can generate professional dialogue with AI voices.", "voice": "aria"}
  ]
}'

教程／讲解

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "Can you walk me through the setup process?", "voice": "jessica"},
    {"text": "Sure. Step one, install the CLI. It takes about thirty seconds.", "voice": "daniel"},
    {"text": "And then what?", "voice": "jessica"},
    {"text": "Step two, run the login command. It opens your browser for authentication.", "voice": "daniel"},
    {"text": "That sounds simple enough.", "voice": "jessica"},
    {"text": "It is. Step three, you are ready to run your first app.", "voice": "daniel"}
  ]
}'

有声读物对话

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "[whispering] Do you hear that?", "voice": "lily"},
    {"text": "[nervously] Hear what? I do not hear anything.", "voice": "harry"},
    {"text": "Exactly. The forest has gone completely silent.", "voice": "lily"},
    {"text": "[worried] That is not a good sign, is it?", "voice": "harry"},
    {"text": "[firmly] We need to move. Now.", "voice": "lily"}
  ]
}'

产品演示

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "So what makes this different from other solutions?", "voice": "brian"},
    {"text": "Three things. Speed, quality, and simplicity.", "voice": "alice"},
    {"text": "That sounds too good to be true.", "voice": "brian"},
    {"text": "[confidently] Let me show you. Watch this.", "voice": "alice"}
  ]
}'

客户支持培训

belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "[frustrated] I have been waiting for twenty minutes and my issue is still not resolved.", "voice": "adam"},
    {"text": "[empathetically] I completely understand your frustration, and I apologize for the wait. Let me look into this right away.", "voice": "sarah"},
    {"text": "I just need my account access restored.", "voice": "adam"},
    {"text": "Of course. I can see the issue here. Give me just a moment and I will have this fixed for you.", "voice": "sarah"},
    {"text": "Okay, thank you.", "voice": "adam"},
    {"text": "[cheerfully] All done! Your access has been restored. Is there anything else I can help with?", "voice": "sarah"}
  ]
}'

提示

1. 变换句子长度 - 将简短回应与较长的解释混合使用
2. 加入回应 - “没错！”，“有意思。”，“嗯。”能让对话更自然
3. 少量使用指导语 - 每轮对话使用一两个即可，不要每行都加
4. 保持片段简短 - 每轮不超过 3 句话，节奏更自然
5. 分配不同语音 - 使用有明显差异的语音，增强辨识度
6. 按口语方式写作 - 使用缩略形式和非正式语言

工作流：对话 + 音乐

# 1. Generate dialogue
belt app run elevenlabs/text-to-dialogue --input '{
  "segments": [
    {"text": "Welcome to the show.", "voice": "george"},
    {"text": "Great to be here.", "voice": "aria"}
  ]
}' > dialogue.json

# 2. Generate background music
belt app run elevenlabs/music --input '{
  "prompt": "Soft podcast background music, non-intrusive",
  "duration_seconds": 30
}' > music.json

# 3. Mix (no inference.sh app mixes audio; use ffmpeg locally)
curl -L -o dialogue.mp3 "<dialogue-url>"
curl -L -o music.mp3 "<music-url>"
ffmpeg -i dialogue.mp3 -stream_loop -1 -i music.mp3 -filter_complex \
  "[1]volume=0.15[bg];[0][bg]amix=inputs=2:duration=first" mixed.mp3

使用场景

• 播客：多主持人节目、访谈
• 有声读物：角色对话场景
• 在线学习：教师与学生的交流
• 讲解视频：问答形式的教程
• 培训：客户服务场景
• 视频脚本：制作前的对话测试

相关技能

# ElevenLabs TTS (single-voice narration)
npx skills add inference-sh/skills@elevenlabs-tts

# ElevenLabs music (background for dialogue)
npx skills add inference-sh/skills@elevenlabs-music

# Dia TTS dialogue (free alternative with S1/S2 tags)
npx skills add inference-sh/skills@dialogue-audio

# Full platform skill (all apps)
npx skills add inference-sh/skills@infsh-cli

浏览所有音频应用：belt app list --category audio