Alvin Lang
Mar 06, 2026 12:19
ElevenLabs releases Voice Design v3, letting customers generate customized AI voices from textual content prompts. The $11B startup continues speedy product growth.
ElevenLabs has launched Voice Design v3, an upgraded device that generates customized AI voices from easy textual content descriptions. Customers describe what they need—age, accent, tone, audio high quality—and the system returns three distinct voice choices inside seconds.
The discharge comes roughly a month after the corporate closed a $500 million Sequence D spherical in February 2026, pushing its valuation to $11 billion. That is a 3x leap from its $3.3 billion Sequence C valuation simply 13 months earlier.
How It Works
The v3 system processes pure language prompts like “middle-aged New Yorker with rising intonation and a half-smile” and outputs three voice candidates. Customers maintain one, discard the remainder, and pay just for the characters of their preview textual content—not per pattern generated.
Two modes can be found: Sensible Voice Design for lifelike performances, and Character Voice Design for fictional characters starting from recreation NPCs to fantasy creatures.
Key enhancements over earlier variations embrace precision audio quality control (customers can specify “excellent audio high quality” or deliberately degraded results like radio static) and what ElevenLabs calls a “smarter prompting engine” educated on expanded accent information.
Market Place
ElevenLabs has moved aggressively since its 2022 founding by ex-Google ML engineer Piotr Dąbkowski and former Palantir strategist Mati Staniszewski. Each Polish natives reportedly began the corporate after rising annoyed with poorly dubbed American movies.
The trajectory has been steep. Sequence A in June 2023 valued the corporate at $100 million. Sequence B eight months later: $1.1 billion. Now they’re sitting at $11 billion with merchandise spanning voice cloning, AI dubbing, and audiobook manufacturing instruments.
A Google Cloud partnership introduced March 2, 2026 expanded their enterprise attain, whereas their audiobook toolkit launched February 9 targets the publishing business’s rising curiosity in AI narration.
Sensible Functions
ElevenLabs positions v3 for recreation builders prototyping character voices, audiobook producers needing distinct narrator types, and localization groups sustaining timing throughout translated dialogue. The corporate additionally highlights accessibility makes use of—constructing branded voices for display screen readers and IVR programs.
The device is dwell now within the ElevenLabs dashboard underneath Voices → My Voices → Add a brand new voice → Voice Design.
For groups burning by voice actor budgets or ready weeks for recording classes, the pitch is simple: kind an outline, click on generate, ship. Whether or not the output high quality matches skilled recordings stays the query enterprise patrons will reply with their wallets.
Picture supply: Shutterstock
