Replaced by Eleven v4 as the top model, though still a solid choice for expressive voices at $0.08 per 1,000 characters.

Kenning's take

What it does

Eleven v3 is an ElevenLabs text-to-speech model that adds audio tags like [whispers] or [sighs] and multi-speaker dialogue across 70+ languages. It launched as an alpha research preview in June 2025 and became generally available in February 2026 with better stability and handling of numbers and symbols. It is available in the ElevenLabs apps and API, and Eleven v4 has since replaced it as the top model.

• Audio tags for emotion and sound
• Multi-speaker dialogue mode
• 70+ languages
• Conversational low-latency variant
• Available in apps and API

Who it suits. Creators and developers making audiobooks, videos and games who want expressive AI voices.

Key facts

Category

AI models

Part of

ElevenLabs (ElevenReader and ElevenLabs app)

Made by

ElevenLabs

Released

3 June 2025

How to get it

in the ElevenLabs app on all plans and in the API

Works on

Web, API

In the UK

Yes

Free plan

10k credits a month on the Free plan

API price

$0.08 per 1,000 characters (v3); v3 Conversational $0.04 per 1,000 characters

Pricing

Plan	Price	Billed
Starter	$6	month
Creator	$22	month
Pro	$99	month
Scale	$299	month
Business	$990	month

source

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.