Eleven v4
Most expressive ElevenLabs speech. 20 voices, 90+ languages, audio tags.
~4.1 credits per 1,000 characters
Run it in Audio Generator- Provider
- ElevenLabs
- Runs in
- Audio Generator
- From
- ~4.1 credits per 1,000 characters
Supported settings
- Up to 9,000 characters
- River · George · Sarah · Daniel · Alice · Brian · Lily · Eric · Matilda · Roger · Bella · Chris · Jessica · Will · Laura · Charlie · Liam · Bill · Callum · Harry
- Up to 300 characters
What Eleven v4 can do
- Text to speech
Made with Eleven v4
Pipelines that run Eleven v4
About Eleven v4
What it does
Eleven v4 is ElevenLabs' newest and most expressive text-to-speech model. In Audio Generator it reads your text in one of 20 ElevenLabs preset voices and returns an MP3. It speaks more than 90 languages, detected from your text, and follows audio tags such as [whispers], [sighs], or [excited] placed where the delivery should change.
When to use it
- Narration, audiobooks, and character voiceovers that need emotional range.
- Scripts in less common languages: every voice speaks the language of the text fluently.
- Lines with laughter, sighs, whispers, or a change of tone partway through.
What to know before you run it
Use the voice instructions for tone, pacing, emotion, or accent. With Enhance prompt on, Pipe2 adds audio tags to your text without changing its words and can pick a voice for you. With it off, your text is read as written and the instructions become one tag at the start. Ellipses add pauses and capital letters add emphasis; SSML is not supported. Text can be up to 9,000 characters. The price grows with the characters sent, audio tags included, and with Enhance prompt on it also includes the enhancement step. Voice cloning, multi-speaker dialogue, and speed control are not available here.
Also known as
You may see this model listed as ElevenLabs v4 or by its API name eleven_v4.