Audio generation
Available now

Eleven v4

by ElevenLabs

ElevenLabs' most expressive voice model.
Audio tags, 90+ languages, and a faster Turbo tier.

Eleven v4

Key features

Hear the range

Cranes in silhouette in a misty marsh at first light
Nature narrationNarration

Calm documentary pace, with room to breathe.

0:00
0:12
A brass key on a desk in a candlelit stone chamber, one chair pulled out
Villain monologueCharacter

A low, gravelly read with a chuckle on cue.

0:00
0:12
A parked car at night with keys in the ignition and rain on the windshield
Two-hander sceneDialogue

Two speakers who answer each other.

0:00
0:12
A mossy cedar forest path in morning mist with a songbird on a branch
Japanese narrationMultilingual

One voice, speaking Japanese like a local.

0:00
0:12

Technical specifications

90+

A native accent in each, from one voice

3,000 chars

Per speech generation on Morphic

2 to 10

Speakers in one scene, 2,000 characters in total

2

Eleven v4 and Eleven v4 Turbo

Use cases

A long ribbon of warm amber light unspooling slowly into deep black

Audiobooks and long-form

A narrator who keeps the same voice from chapter one to the last page, with the pacing to carry a slow scene.

A sharp burst of white-gold light radiating from a single bright point

Ads and promos

Thirty seconds that have to sell. Direct the energy with a tag and regenerate the one line that isn't landing.

Three translucent glass masks floating in haze, each lit from a different side

Characters and games

Whispers, laughs, a sarcastic aside. The performance range for villains, sidekicks and every voice in the village.

Two ribbons of light, one amber and one cool silver, answering each other across the dark

Podcasts and audio drama

Script the whole scene with every speaker in it. The replies react to each other, so the exchange sounds like people talking.

A clean series of luminous steps rising and receding into depth

Explainers and courses

Clear, even narration for lessons and product tours. Change one step in the script and regenerate just that line.

A single beam of light splitting into parallel beams that stay in step

Localized voice-over

Run the same script in Spanish, Japanese or Portuguese and keep one voice, now with the accent each market expects.

Prompt examples

A dim late-night cafe with a coat hanging on the door and rain on the window

Whispered warning

[whispers] Don't turn around. [pause] He's been by the door all night.

0:00
0:12
Edit prompt
A chocolate cake with three slices already gone and a fork on the tablecloth

Laugh mid-line

I said the cake was sugar-free. [laughs] She'd had three slices by then.

0:00
0:12
Edit prompt
An empty underground platform with amber edge lights and a train glowing in the tunnel

Written direction

[low, steady, urgent] Everyone off the platform. Walk, don't run.

0:00
0:12
Edit prompt
A smartphone lighting up on a wooden desk beside an open notebook

Sound in the take

[phone buzzing] That's work again. [sighs] I'll call them back later.

0:00
0:12
Edit prompt
A small ferry tied to a snowy dock on an iced-over river at dawn

Audiobook opening

Chapter three. The river froze overnight... and the ferry never left.

0:00
0:12
Edit prompt
A rustic homemade loaf on a floured board in a sunny farmhouse kitchen

Spanish, native accent

[warm] Bienvenidos. Hoy preparamos el pan de la abuela, paso a paso.

0:00
0:12
Edit prompt

Overview

Eleven v4 is ElevenLabs' newest text-to-speech model, released on September 28, 2026 and the default for Speech on Morphic. It reads tone and pacing from the script, takes audio tags for delivery and sound effects, runs multi-speaker dialogue, and speaks 90+ languages with a native accent.

Eleven v4 on Morphic

Open Speech and ElevenLabs is already selected on Eleven v4; choose ElevenLabs Turbo for the faster tier. Paste a script of up to 3,000 characters with audio tags inline, or switch to Dialogue for two to ten speakers. For the full tag list and how it compares with Eleven v3 and Turbo, see the Eleven v4 prompt guide.

Simple pricing

Get started for free today, with the option to upgrade or cancel anytime.

Basic

$19/ month
billed as $0 per year

2400 monthly credits

1 user only

All models

Workflows

Standard

$24/ month
billed as $0 per year

3625 monthly credits

1 user only

All models

Workflows

Pro

$45/ month
billed as $0 per year

6350 shared monthly credits

1 user

+ up to 4 more at extra cost

All models

Workflows

Max

$170/ month
billed as $0 per year

24650 shared monthly credits

1 user

+ up to 9 more at extra cost

All models

Workflows

Enterprise

For higher limits

Custom

pricing and billing terms

High-volume credits
Custom seat limits
All models
Workflows
Pricing Gradient

Free

For playing around

$0

forever free

Up to 20 credits
1 user only
Limited models
Workflows
Compare plan details

FAQs

What is Eleven v4?
Eleven v4 is ElevenLabs' text-to-speech model released on September 28, 2026, and the model behind ElevenLabs speech and dialogue on Morphic. It reads tone and context from a script, takes audio tags for delivery and sound effects, performs multi-speaker dialogue, and speaks more than ninety languages.
What is the difference between Eleven v4 and Eleven v3?
Eleven v4 performs more from context, keeps a voice steadier across retakes and long pieces, follows audio tags more reliably, and adds more than twenty languages. Two things work differently: the speed and style controls are gone, and a voice speaking a language other than its own now uses a native accent for that language.
What is the difference between Eleven v4 and Eleven v4 Turbo?
They share the same voices and the same stability and resemblance settings. Eleven v4 gives the fullest performance and also runs multi-speaker dialogue. Eleven v4 Turbo is the faster tier for speech only, which suits drafts, quick retakes and high-volume lines. On Morphic, pick ElevenLabs for v4 or ElevenLabs Turbo for Turbo.
How many languages does Eleven v4 support?
More than ninety, according to ElevenLabs, up from over seventy on Eleven v3. Any voice can speak any of them. Delivery is more fluent in some languages than others, so test a line in your target language before generating a full script.
Which audio tags does Eleven v4 understand?
Tags are written in square brackets and are free text rather than a fixed list. Common ones cover emotion and delivery ([excited], [sarcastic], [whispers]), reactions ([laughs], [sighs], [exhales]), pauses ([pause], [long pause]) and sound effects ([applause], [phone buzzing]). You can also write a direction in your own words, such as [dry, quietly pleased].
How do I add a pause or change the speed in Eleven v4?
Eleven v4 has no speed control and does not read SSML break tags. Use [pause] or [long pause] for a gap, an ellipsis for a softer beat, and a delivery tag such as [rushed] or [slowly] to change the pace of a line. Punctuation and capital letters for emphasis shape the rhythm too.
Why does my voice have a different accent in another language?
That is how Eleven v4 is built. When a voice speaks a language other than the one it was recorded in, it uses a native accent for the new language, so a British voice speaking Spanish sounds like a Spanish speaker. An accent tag such as [strong French accent] can steer it, though results vary.
Is Eleven v4 free to use on Morphic?
Yes, you can start on Morphic's free tier with no payment. Eleven v4 is the default speech model, so it is already selected when you open Speech. Paid plans add more credits for longer scripts and commercial volume.

More about Eleven v4

Eleven v4 prompt guide

How to use Eleven v4: audio tags you can hear, pacing without SSML, dialogue scripts, and what changed from Eleven v3 and Turbo.

Other models

Kling 4.0 Flash

Kling

Kuaishou's speed-tuned Kling 4.0 for high-volume video. Fast 3 to 20 second clips at 720p.

Kling 4.0

Kling

Kuaishou's next Kling model, out in October. 30-second clips and up to 10 keyframes.

ChatGPT Images 2.5

OpenAI

OpenAI's image model, released September 2026. Sharper detail, precise edits, faster.

Lyria 3.5

Google DeepMind

Google's best-sounding music model. Full songs with structure, vocals, and lyrics.