Eleven v3
Eleven v3 is ElevenLabs' most expressive text-to-speech model, released as an alpha on 3 June 2025 and to general availability on 14 March 2026. Its signature is audio tags, stage directions in square brackets such as whispers or sighs that the model acts, plus a dialogue mode for multi-speaker scripts and more than 70 languages. It is not a real-time model; ElevenLabs points conversational use at Flash. On fal and on ElevenLabs' own API it costs $0.10 per thousand characters, and the free plan includes it for personal use.
Anvisha Pai, Co-founder & CEO, Voyager
ElevenLabs · released
›What we asked for
Hi, I'm Eleven v3, the voice model from ElevenLabs. I read scripts in more than seventy languages, and if you write a stage direction in square brackets, I'll try to act it. Here is the same script every voice on this site reads, so you can compare us.
fal-ai/elevenlabs/tts/eleven-v3
voice: Sarah, stability: 0.5
returned 16.1 seconds of mp3
At a glance
- Price
- $0.10 per thousand characters on fal and on ElevenLabs' API; our four scripts cost about ten cents in total.
- Voices
- Twenty named stock voices on fal, from Aria to Bill; we used Sarah. Cloning lives on ElevenLabs' own platform.
- Audio tags
- Square-bracket directions like [whispers] that the model performs; our scripts used none, to keep every model on the same words.
- Languages
- More than 70.
- Not real-time
- A larger, slower model; ElevenLabs recommends Flash for conversational latency.
Anvisha's take
We have not scored the four readings by ear; press play and judge the voice yourself. What we measured: Eleven v3 read the four scripts in 13 to 25 seconds of audio each, taking seven to ten seconds to return, at $0.10 per thousand characters, about a dime for all four. On the narration script it read the same 385 characters in 24.5 seconds, between Kokoro's 22.9 and the 25 to 26 of MiniMax and Gemini. The stock Sarah voice was used with no audio tags, so the emotion script is a test of what the model does from the words alone.
Eleven v3 real output: the same 4 scripts every model gets
Every voice model on this site gets the same 4 scripts, so you can compare like with like. These are the readings Eleven v3 returned, one try each, nothing re-rolled or retouched. Generated on . The 4 readings took 7 to 10 seconds each and cost $0.0976 in total.
Pacing, warmth and whether the pauses land where a voice actor would put them.
›What we asked for
Meet the Ember mug. It keeps your coffee at exactly the temperature you choose, for up to ninety minutes, and it tells your phone when it's ready. Warm from the first sip to the last. Ember. Coffee, on your terms.
fal-ai/elevenlabs/tts/eleven-v3
voice: Sarah, stability: 0.5
returned 13.8 seconds of mp3
Steady narration over a long paragraph, clean list rhythm and a natural close.
›What we asked for
A pour-over works in four steps. First, rinse the paper filter with hot water so it doesn't taste of paper. Second, add the grounds and pour just enough water to wet them, then wait thirty seconds while they bloom. Third, pour the rest of the water in slow circles. Fourth, wait about three minutes for it to drip through. That's it: a cup that tastes like the coffee, not the machine.
fal-ai/elevenlabs/tts/eleven-v3
voice: Sarah, stability: 0.5
returned 24.4 seconds of mp3
Range: excitement, then a hushed aside, from the words alone with no tags.
›What we asked for
We did it! We actually did it! I can't believe it's finally over. Okay. Okay, deep breath. Let's not tell anyone until Monday, alright? Just... let me have this one quiet night.
fal-ai/elevenlabs/tts/eleven-v3
voice: Sarah, stability: 0.5
returned 13.0 seconds of mp3
Reading numbers, times, currency, an email address and two hard names correctly.
›What we asked for
Your order ships on the 24th of September 2026 and arrives by 9:15 a.m. The total is $1,249.99, including VAT. Questions? Ask for Siobhan Nguyen or Ravi Parikh on 0800 555 0199, or email help@moda.app.
fal-ai/elevenlabs/tts/eleven-v3
voice: Sarah, stability: 0.5
returned 24.7 seconds of mp3
How the cost under each picture was worked out: fal's Eleven v3 rate of $0.10 per 1,000 characters (fal model page, read 2026-09-10) for the 213 characters sent; fal does not return the charge.
Eleven v3 pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | 1,000 characters | $0.10 | fal model page, read |
| ElevenLabs API | 1,000 characters v3 and Multilingual v2 at $0.10; v3 Conversational at $0.05. | $0.10 | ElevenLabs API pricing, read |
Is Eleven v3 free?
ElevenLabs' free plan gives 10,000 credits a month, roughly 10,000 characters, with Eleven v3 included, for non-commercial use. Commercial rights start at the $6 Starter plan with 30,000 credits; Creator is $22 for 121,000 and adds professional voice cloning.
- •Free: 10,000 credits a month, non-commercial, v3 included.
- •Starter $6 a month: 30,000 credits and a commercial licence.
- •Creator $22 a month: 121,000 credits and professional voice cloning.
- •fal: per character, no free tier.
ElevenLabs pricing, read
Where to use Eleven v3
- ElevenLabs · Free plan and paid tiers; voice cloning, dubbing and agents.
- ElevenLabs API · Model id eleven_v3.
- fal · $0.10 per thousand characters, stock voices.
- Voyager · Runs inside Voyager, the agent for creative work. Private preview by waitlist.
Voyager
Run Eleven v3 inside Voyager
Voyager is an agent for creative work that runs every model in this reference, Eleven v3 included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running Eleven v3
- •24.5 seconds for the paragraph: quicker than MiniMax at 25.2 and Gemini at 25.8, slower than Kokoro at 22.9, on the same 385 characters. See the explainer paragraph reading.
- •No audio tags were used, so the excitement and the hushed aside come from the words; v3's tags are the feature this fixture deliberately does not exercise. See the emotional line reading.
- •Returned in seven to ten seconds per reading; ElevenLabs points real-time use at Flash for a reason.
Eleven v3 limits and API parameters
| Limit | Value |
|---|---|
| Latency | Not real-time; our readings took seven to ten seconds to return. Inworld on Eleven v3 GA, read |
| Voices on fal | Twenty stock voices; no cloning through fal. fal API reference, read |
›Every setting, for developers
fal (fal-ai/elevenlabs/tts/eleven-v3)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| text | string | The script, with optional [audio tags] Required. | |
| voice | string | Aria, Roger, Sarah, Laura, Charlie, George, Callum, River, Liam, Charlotte, Alice, Matilda, Will, Jessica, Eric, Chris, Brian, Daniel, Lily, Bill | Rachel |
| stability | number | 0 to 1 | 0.5 |
| language_code | string | ISO 639-1 code | |
| timestamps | boolean | true, false Word-level timing for subtitles. | false |
| apply_text_normalization | enum | auto, on, off | auto |
What Eleven v3 will not say
ElevenLabs' terms bar cloning a voice without consent, impersonation, fraud, harassment and sexual content involving minors, and the free plan is non-commercial with attribution. Cloned voices need the speaker's permission; ElevenLabs runs a voice-captcha check for professional clones.
Alternatives to Eleven v3
Frequently asked questions
What is Eleven v3?
Eleven v3 is ElevenLabs' expressive text-to-speech model, generally available since 14 March 2026, with audio tags for emotion, a dialogue mode and more than 70 languages.
How much does Eleven v3 cost?
$0.10 per thousand characters on fal and on ElevenLabs' API, read on 10 September 2026. ElevenLabs' plans bundle credits, from a free 10,000 a month.
Is Eleven v3 free?
On ElevenLabs' free plan, yes, with 10,000 credits a month for non-commercial use. Commercial use starts at $6 a month.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
