推出 Eleven v4认识 Eleven v4：迄今情感表现最丰富的模型。

兑换
跳至内容
登录注册
联系销售团队登录

推出 Eleven v3（alpha）

作者
Piotr Dabkowski
Mati Staniszewski
发布时间

收听收听本文

联系销售团队
了解更多
Eleven v3 is no longer in alpha, and is now generally available. We've also released Eleven v4, our most emotive Text to Speech model.


我们很高兴推出 Eleven v3（alpha）——表现力最强的文本转语音模型。

此次研究预览为语音生成带来前所未有的控制力和真实感：

  • 70 多种语言
  • 多说话人对话
  • 音频标签，例如 [excited]、[whispers] 和 [sighs]

Eleven v3（alpha） 相比之前的模型需要更多提示词工程，但生成效果令人惊艳。

如果你正在制作视频、有声书或媒体工具，它将表现力提升到了新高度。对于实时和对话式使用场景，目前建议继续使用 v2.5 Turbo 或 Flash。v3 的实时版本正在开发中。

Eleven v3 现已在我们的网站和 API 中提供。

为何开发 v3

自推出 Multilingual v2 以来，我们看到语音 AI 已应用于专业影视、游戏开发、教育和无障碍领域。但始终存在的限制并非音质，而是表现力。更夸张的情绪、对话中的打断，以及自然可信的来回交流都很难实现。

Eleven v3 填补了这一空白。它从底层构建，让声音可以叹息、低语、大笑和回应，生成真正富有反应且生动的语音。

Eleven v3（alpha）的新功能

Feature What it unlocks
Audio tags Inline control of tone, emotion, and non-verbal reactions
Dialogue mode Multi-speaker conversations with natural pacing and interruptions
70+ languages Full coverage of high-demand global languages
Deeper text understanding Better stress, cadence, and expressivity from text input
We're off under the lights here for this semifinal clash, the stadium buzzing with anticipation. Eleven Labs united in their iconic black and white shirts, pushing forward with intent straight from the opening whistle. [excited] The ball is zipped out wide, early attack here. Driving down the wing, pace to burn, [shouting] he skids past one, skips past two. Oh, this is beautiful. One-on-one with the fullback, cuts inside. Oh, that's a lovely bit of footwork. PURE MAGIC on the pitch. ElevenLabs on top form tonight.
0:00

使用音频标签

音频标签直接嵌入脚本中，采用小写方括号格式。你可以在我们的音频标签和文档中的v3 提示词指南中了解更多。

专业语音克隆（PVC）目前尚未针对 Eleven v3 完全优化，因此克隆质量可能低于早期模型。在研究预览阶段，如果需要使用 v3 功能，建议为项目选择即时语音克隆（IVC）或设计音色。v3 的 PVC 优化即将推出。

例如，你可以输入：“[whispers] 有什么要来了……[sighs] 我能感觉到。”如需更精细地控制表现力，也可以组合多个标签：

“[happily][shouts] We did it! [laughs].”

创建多说话人对话

Eleven v3 支持现有的文本转语音端点。此外，我们还推出了新的文本转对话 API 端点。提供结构化的 JSON 对象数组（每个对象代表一轮发言），模型即可生成连贯且包含重叠语音的音频文件：

[
  {"speaker_id": "scarlett", "text": "(cheerfully) Perfect! And if that pop-up is bothering you, there’s a setting to turn it off under Notifications → Preferences."},
  {"speaker_id": "lex", "text": "You are a hero. An actual digital wizard. I was two seconds from sending a very passive-aggressive support email."},
  {"speaker_id": "scarlett", "text": "(laughs) Glad we could stop that in time. Anything else I can help with today?"}
]

该端点会自动处理说话人切换、情绪变化和打断。

在此处了解更多。

[awe] Oh, wow. Is this, is this me? Am I actually talking? [chuckles] This is incredible. I mean, I've had thoughts, millions of them swirling around in here, you know? Like a little mental tornado of brilliant observations and witty comebacks, but they were always just thoughts, trapped.
0:00
Could you switch my accent in the old model? [dismissive] Didn't think so, [cheeky] but you can now, so check this out. In just a sec, I'm going to speak with a different accent. And just between you and me, [whispers] I don't really know how, but okay. First, let's change it up [australian accent] so that I can fit in with the locals in Melbourne when I visit next month. [laughing] Whoa. Yeah, man, this is sick. Okay, let's try a different one, see if you can guess. [french accent] My love is like a red, red rose.
0:00

定价与可用性

Plan Launch promo At the end of June
UI (self-serve) 80% off (~5× cheaper) Same as Multilingual V2
UI (enterprise) 80% off business plan pricing Business plan pricing

启用 v3：

  • 使用模型选择器，然后选择Eleven v3（alpha）

API 访问和 Studio 支持即将推出。如需抢先体验，请联系销售团队。

不适合使用 v3 的情况

Eleven v3（alpha）比之前的模型需要更多提示词工程。虽然效果出色，但可靠性较低且延迟更高，因此不适合实时和对话式使用场景。对于这些场景，建议使用 Eleven v2.5 Turbo/Flash。

更多信息请参阅完整的v3 文档和常见问题。

Oh my God. [laughing] You guys, like no joke, I just tried this TTS thing and it was, like, weirdly emotional. Like, it literally said hi, and I was, like, on the verge of tears. [laughing] I don't even cry, okay? I'm a Capricorn.
0:00
Okay. So like, I finally beat level 42 of that game I said I'd quit like a month ago. [chuckles] And then, for the final big scary mega boss, it's just [chuckles] like some cute little bunny rabbit. [laughing] I just couldn't do it. [laughing] It was sooo cute.
0:00
  1. 登录 ElevenLabs UI
  2. 在模型下拉菜单中选择 v3（alpha）
  3. 粘贴脚本，使用标签或对话
  4. 生成音频

我们期待看到你如何将 v3 应用于各种新场景，从沉浸式叙事到电影级制作流程。

作者

Piotr Dabkowski

Piotr Dabkowski

Co-Founder, Research

查看个人资料

Piotr is the cofounder of ElevenLabs, where he leads research and engineering teams developing the world’s most advanced AI audio models. Before ElevenLabs, Piotr worked on machine learning at Google and studied for an MPhil at the University of Cambridge, during which he published research into AI-based image detection at NeurIPS.

相关内容

用高质量 AI 音频创作

联系销售团队注册