How to describe a voice in text and get back the one you hear in your head. Based on the ElevenLabs Voice Design docs and the product page, plus what works in practice.
First, create an account: 👉 https://elevenlabs.io/creative
Voice Design is available on the free plan. Saving a voice takes one voice slot, and the number of slots depends on your plan. Check the pricing page for slot counts.
Before you design a voice, check whether it already exists. The Voice Library has more than 20,000 voices, filterable by accent, age, gender, and use case: 👉 https://elevenlabs.io/app/voice-library
If nothing fits, continue with the guide below.
Voice Design builds a new voice from a written description. You type what the voice should sound like, the model generates three previews, you pick one and save it to My Voices. From there it works like any other voice in Text to Speech, Studio, Flows, or the API.
Use it when the Voice Library does not have the voice you need. If you need a very specific character, an exact age and accent combination, or a voice for a fictional creature, Voice Design fills that gap.
Two things to know before you start: