落笔之前，先听为敬
本页所有模型的真实生成结果。每张卡片都展示了对应的文案或提示词。
Hey, take a deep breath. You’ve been doing great — even when it doesn’t feel like it. Sometimes, the smallest step forward is still progress.
Hello，welcome to WaveSpeedAI！
High-energy cyberpunk synthwave, driving analog bassline, retro futuristic synthesizers, punchy electronic drums, neon noir atmosphere, 120 BPM, mechanical textures.
I don’t need luck. I need focus. Every move, every second — it all leads to one goal.
Hello, little friend! I am your learning partner, Dr. Bot. Today, we will explore the amazing world of numbers together. Are you ready for a fun adventure? Let's go!
Modern deep house fashion runway music, stylish and elegant, groovy bass, rhythmic hi-hats, vocal chops, luxury brand advertisement vibe, confident and cool.
Meet the future of creativity — powered by AI, designed for humans. This is innovation that speaks in your voice.
Some things define eternity without a single word. They are not a product of time, but a witness to it. This is a classic, born for you—a perfect fusion of craftsmanship and art.
Steampunk ambient atmosphere, ticking clock sounds, mechanical gears clicking, steam hissing, soft acoustic guitar in the background, mysterious and studious.
In a world built on shadows and steel, only one voice can bring the story to life — yours.
Look at this! A golden, crispy crust wrapped around a tender, juicy center—the aroma is simply irresistible. Just one bite, and that rich flavor will melt on your tongue. This is absolutely the best dish I have ever tasted!
Once upon a time, in a forest painted with golden leaves, a little fox discovered that courage often comes wrapped in kindness.
从第一行文字到完整配乐，一站搞定
Welcome to our advanced text-to-speech system! Experience high-quality voice synthesis with natural pronunciation and clear articulation.
2. 用 Multilingual v2 说更多语言
多语言模型可用相同的声音与参数朗读英语以外的文案。
生成Oh, brilliant. Another 'world-changing' idea, I'm sure. I can barely contain my excitement, really. So, I guess we're all just supposed to start working on this right now, right?
3. 用 Eleven Music 为瞬间配乐
描述曲风、情绪、乐器和节奏。设置最长五分钟的时长并保持纯乐器演奏,或在提示词中写入歌词以生成人声。
生成Hard Trap, Hip Hop, 808 bass, energetic flow, confident male rap vocals, brass stabs, hype music, motivational sports anthem. [Verse] Laced up tight, ready to go Putting on a major show Sweat and tears on the floor Coming back to get some more [Chorus] Eyes on the prize, reach for the net The greatest game you ever met I take the shot, I make the play Doing this every single day [Outro] Nothing but net. Yeah, we winning.
4. 调整朗读方式
Stability 与 Similarity 滑块，加上 Speaker boost 开关，无需改写文案即可改变每一句的朗读方式。
5. 或接入你的产品
每个模型都有独立的 API 端点，输入与此处相同，你应用中的文案一次请求即可变成音频。
查看 APIPOST /api/v3/elevenlabs/eleven-v3 { "text": "Every setback is fuel.", "voice_id": "Talia", "stability": 0.5 }
模型与定价
价格按每次生成计算并实时更新。分辨率、时长或质量等选项可能改变最终价格，生成前可在生成器中查看。
| 模型 | 输入 | 起价 | |
|---|---|---|---|
| Eleven v3 Text to Speech | 文本 | $0.20 | 生成 |
| Eleven Multilingual v2 Text to Speech | 文本 | $0.20 | 生成 |
| Eleven Music | 提示词 | $0.10 | 生成 |
常见问题
该从哪个 ElevenLabs 语音模型开始?
从 Eleven v3 开始，它是这里最新的语音模型。当你想用另一种语言的文案比较声音效果时，再切换到 Multilingual v2；两者接受相同的文本、声音和参数。
如何选择声音?
在音频生成器的 Voice 字段中输入声音名称，例如 Talia、Darian 或 Elowen。调用 API 时，同样的名称填入 voice_id 字段。
Stability、Similarity 和 Speaker boost 分别是什么?
它们都是 ElevenLabs 的语音参数。Stability 和 Similarity 取值范围都是 0 到 1，分别影响朗读的稳定性以及与所选声音的相似程度；Speaker boost 是一个开关。在同一句文案上尝试几个不同的值，就能听出区别。
Eleven Music 生成的音乐最长可以多久?
时长在生成器中设置，从 5 秒到 5 分钟不等。默认为纯乐器演奏；关闭该选项并在提示词中写入歌词，即可生成带人声的版本。
可以在自己的代码中调用这些模型吗?
可以。表格中的每个模型都有对应的 API 页面和请求示例，文档中也说明了身份验证和结果获取方式。