New model

STORM Ensemble 3.0

STORM Ensemble 3.0 speaks 64 African and Nigerian languages, plus English, through 166 named voices. Every voice in the film below is the model, with nothing edited.

64languages
166named voices
3Bparameters
24 kHzstudio-grade audio
Voices from the film

One welcome, sixteen voices.

Each clip is one of the model's named speakers saying “Welcome, we are glad you are here” in their own language.

OluwaseunYoruba

HadizaHausa

ChidiIgbo

BlessingNigerian Pidgin

JumaSwahili

NomvulaZulu

FatimaArabic

AkosuaAkan

AxmedSomali

KhadyWolof

HanitraMalagasy

LoratoTswana

KwameEwe

MagrietAfrikaans

TapiwaShona

ThaboXhosa

Why it matters

One model for a continent of voices.

Build once and speak to customers across Africa, without stitching together a different system for every language.

/01

One model, every language

Sixty-four languages and English from a single model, so quality and integration stay the same wherever you launch.

/02

Voices you call by name

Pick from 166 named voices. Each one is ready to use, with no setup or training on your side.

/03

Natural in every language

Pronunciation, tone and rhythm that native speakers recognise as their own.

/04

Studio-quality audio

Clear 24 kHz speech, ready for apps, calls, media and broadcast.

Quality first

Every language was evaluated speaker by speaker before release. Voices that did not meet the bar were left out.

Put STORM Ensemble 3.0 to work.

Tell us the languages and the voices you need, and we will set you up.