Sign up and save up to 50% on a yearly plan.Sign up free →Black Friday: up to 50% off yearly plans.See the deal →

ElevenLabs' best voice, ready for your video.

ElevenLabs v4 AI voice generator

Paste a script up to 10,000 characters, pick one of 20 voices and get an MP3 voiceover. Audio tags direct the delivery, in 90+ languages.

From $0.068 per 1,000 characters on MAX yearly

  • 90+ languages
  • 20 voices
  • Up to 10,000 characters a read

● Made with ElevenLabs v4

Hear ElevenLabs v4

Every sample is one read, with the script that made it and what it cost.

  • “Here is how it works. Paste the link to your product, pick a format, and in a few minutes you have a finished video ad, with captions, music and a voiceover.”

    Aria
    157 characters · 3 credits
  • “[excited] Ladies and gentlemen, the doors are open! [whispering] But between us, the best seats went in the first five minutes. [laughing] Better luck next time.”

    Daniel · audio tags
    161 characters · 3 credits
  • “Built for the long haul. A steel frame, a lifetime warranty and a battery that lasts the whole week. The new Ridgeline e-bike. Ride further.”

    Brian
    140 characters · 3 credits
  • “The lighthouse keeper counted the ships every night. Forty years, and not one of them had ever stopped. Until the night the fog rolled in and a small boat turned toward the shore.”

    Lily
    179 characters · 3 credits

● Why people pick it

What is ElevenLabs v4 good at?

01

Highest quality

ElevenLabs' newest generation and its highest-quality model, with better voice accuracy and consistency than v3.

02

90+ languages

Reads more than ninety languages, and keeps a voice fluent in the target language instead of carrying its accent over.

03

Audio tags

[whispering], [shouting], [laughing]: direct the delivery from the script, with more nuance than earlier models.

04

Long scripts

Up to 10,000 characters a read, about ten minutes of speech: twice what v3 takes.

● Use cases

What people make with ElevenLabs v4

Video narration

ElevenLabs' own model guide names v4 for professional content and video narration.

Audiobooks

Long-form narration with complex emotional delivery, at up to 10,000 characters a read.

Multilingual campaigns

The same ad in many languages, with the voice kept consistent across the switch.

● Side by side

How does ElevenLabs v4 compare?

The other voice models on OpenClips, on the same credits.

SpecElevenLabs v4ElevenLabs v3ElevenLabs v4 TurboMiniMax Speech 2.8 Turbo
Made byElevenLabs v4ElevenLabsElevenLabs v3ElevenLabsElevenLabs v4 TurboElevenLabsMiniMax Speech 2.8 TurboMiniMax
Price per 1,000 charactersElevenLabs v416.3 creditsElevenLabs v316.3 creditsElevenLabs v4 Turbo8.2 creditsMiniMax Speech 2.8 Turbo12.3 credits
A 1,000-character readElevenLabs v417 creditsElevenLabs v317 creditsElevenLabs v4 Turbo9 creditsMiniMax Speech 2.8 Turbo13 credits
Longest scriptElevenLabs v410,000 charactersElevenLabs v35,000 charactersElevenLabs v4 Turbo10,000 charactersMiniMax Speech 2.8 Turbo5,000 characters
VoicesElevenLabs v420ElevenLabs v320ElevenLabs v4 Turbo20MiniMax Speech 2.8 TurboBy voice id
OutputElevenLabs v4MP3ElevenLabs v3RASTERElevenLabs v4 TurboRASTERMiniMax Speech 2.8 TurboRASTER

Prices are the catalog's, quoted before every render.

● Specs

ElevenLabs v4 specs

StatusLive in the app, the API and the MCP
Made byElevenLabs
ModalityText to speech
InputsText script
Longest script10,000 characters
Voices20: Aria · Roger · Sarah · Laura · Charlie · George · Callum · River · Liam · Charlotte · Alice · Matilda · Will · Jessica · Eric · Chris · Brian · Daniel · Lily · Bill
Languages90+ languages
Languagemixed · af · ar · hy · as · ast · az · be · bn · bs · bg · my · yue · ca · ceb · hr · cs · da · nl · en · et · fil · fi · fr · gl · ka · de · el · gu · ha · he · hi · hu · is · id · ga · it · ja · jv · kn · kk · ko · ky · lv · ln · lt · lb · mk · ms · ml · mt · zh · mi · mr · mn · ne · no · oc · or · ps · fa · pl · pt · pa · ro · ru · sr · sd · sk · sl · so · es · sw · sv · tg · ta · te · th · tr · uk · ur · uz · vi · cy · yo
Output formatMP3 22.05 kHz 32 kbps · MP3 44.1 kHz 32 kbps · MP3 44.1 kHz 64 kbps · MP3 44.1 kHz 96 kbps · MP3 44.1 kHz 128 kbps · MP3 44.1 kHz 192 kbps
Similarity boost0–1
Stability0–1
Audio tags[whispering] [shouting] [laughing] and more, the same tags as v3
PronunciationIPA between slashes in the script
SSMLNot supported
OutputMP3 audio
Price16.3 credits per 1,000 characters · $0.068 per 1,000 characters on MAX yearly
A 1,000-character read17 credits

● How it works

How do you use ElevenLabs v4 on OpenClips?

01

Describe it

Write the script, up to 10,000 characters. Every word is read aloud.

02

Pick a voice

20 voices. Play a sample, then keep the one that fits the script.

03

Pick the settings and generate

Choose the settings ElevenLabs v4 offers, then generate. The result lands in your library.

● Tips

Tips for better results with ElevenLabs v4

  • Pin the pronunciation

    Write a brand name in IPA between forward slashes and v4 reads it exactly. No XML needed.

  • Two sliders, not four

    v4 uses stability and similarity only. For a slower or more stylised read, use tags and punctuation.

  • Set the language

    A language code helps when the script mixes languages or a name could be read either way.

● Made for

What can you make with ElevenLabs v4?

Video narrationAudiobooksExplainersProduct demosE-learningGame charactersMultilingual adsPodcast intros

● Questions

ElevenLabs v4 FAQ

ElevenLabs v4 is a text-to-speech model by ElevenLabs. ElevenLabs' newest and highest-quality text-to-speech model.

ElevenLabs calls v4 a net upgrade over v3, with better results in almost every case: voice accuracy, consistency, emotion, delivery, tags and languages. It also notes you may still prefer how a particular voice sounded on v3.

Yes. Tags like [whispering], [shouting] and [laughing] go in square brackets in the script, and the same tags work on v3 and v4.

v4 is the highest-quality model. v4 Turbo is built for real-time use at about 100 ms median latency and costs about half as much per character on OpenClips. For a finished voiceover, pick v4.

ElevenLabs v4 opens with your first purchase: any plan or credit pack. A new account still gets ⚡30 on signup with no card, for the models the free plan runs.

ElevenLabs v4 costs 16.3 credits per 1,000 characters, paid from your OpenClips credits. That is from $0.068 per 1,000 characters on MAX yearly. A 1,000-character script costs 17 credits. The charge is the characters you send, rounded up once.

Up to 10,000 characters a read. A longer script is refused before any credit is held, so split a long script into reads.

20 voices: Aria, Roger, Sarah, Laura, Charlie, George, Callum, River, Liam, Charlotte, Alice, Matilda, Will, Jessica, Eric, Chris, Brian, Daniel, Lily, Bill. Pick one per read.

90+ languages, as ElevenLabs documents it.

An MP3 file, saved to your library, ready to download or to drop on a video.

Per 1,000 characters: ElevenLabs v4 17 credits, ElevenLabs v3 17 credits, ElevenLabs v4 Turbo 9 credits, MiniMax Speech 2.8 Turbo 13 credits. The table above sets their scripts, voices and files side by side.

Yes. ElevenLabs v4 runs from the OpenClips app, the REST API and the MCP server, with the same settings and the same price.

● The receipts

Where do these ElevenLabs v4 facts come from?

Every link goes to the vendor's own page.

Voice your next video with ElevenLabs v4

90+ languages, 20 voices, 10,000 characters a read.

Generate with ElevenLabs v4

What ElevenLabs v4 does

ElevenLabs v4 is the first model in ElevenLabs' newest generation and its highest-quality text to speech. Write a script up to 10,000 characters, pick a voice, and the read lands in your library as an MP3.

Why it sounds better than v3

ElevenLabs reports better voice accuracy, consistency, emotion and delivery than ElevenLabs v3. It also handles accents better: when a voice reads another language, v4 speaks it fluently instead of carrying the voice's own accent across.

ElevenLabs v4 vs v4 Turbo

ElevenLabs v4 Turbo is the real-time version, built for voice agents at about 100 ms median latency. It costs about half as much per character. When the read is the final voiceover, v4 is the one to pick.