SAVRN
Search Contact SAVRN

Organization

Silencio Voice AI

SilencioNetwork

Speech data for underrepresented languages, accents and niche domains, at scale. 2.5M+ consented contributors | 180+ countries | ~350 languages collectable | ~500,000 hours off the shelf Proprietary, first-party recordings with consent and provenance records, not available in any other dataset. Human-validated transcription for every language, tailored to your needs. Anything not off the shelf…

Models in Library0
Datasets in Library3
Models on Hugging Face—
Followers11

Datasets

Dataset · Audio classification

english-accents-speech

Silencio Voice AI

English from 282 speakers born in 35 countries. 513 clips, 9.3 hours, each labelled with the speaker's country of birth, first language, self-reported accent or regional variety, declared English proficiency, age band, gender and recording device. Audio and speaker metadata only. Human-validated transcription is available on request. English as it is actually spoken worldwide is mostly not American or British. Across Silencio's English catalogue, 59% of recorded hours come from speakers born in Africa and only about 11% from the US, UK, Canada, Australia, Ireland and New Zealand combined. Nigeria, the largest single origin, is 16% of hours. This sample covers 35 countries of birth.…

Publicly accessible cc-by-nc-4.0 n<1K

Dataset · Audio classification

spanish-accents-speech

Silencio Voice AI

Spanish from 12 speakers born in 9 countries. 19 clips, 0.3 hours, each labelled with the speaker's country of birth, first language, self-reported accent or regional variety, declared Spanish proficiency, age band, gender and recording device. Audio and speaker metadata only. Human-validated transcription is available on request. Silencio's Spanish catalogue is 65% Latin American by recorded hours, led by Venezuela, with Spain at 9%. It also holds a substantial body of Spanish spoken as a second language: 18% of hours come from speakers born in Africa, notably Nigeria, Kenya and Egypt. This sample covers 9 countries of birth across Latin America, Spain, Africa, the Middle East and the…

Publicly accessible cc-by-nc-4.0 n<1K

Dataset · Audio classification

french-accents-speech

Silencio Voice AI

French from 20 speakers born in 9 countries. 32 clips, 0.5 hours, each labelled with the speaker's country of birth, first language, self-reported accent or regional variety, declared French proficiency, age band, gender and recording device. Audio and speaker metadata only. Human-validated transcription is available on request. Most French speakers today live in Africa, and Silencio's French catalogue reflects that: 74% of recorded hours come from speakers born in Africa, led by Benin, Senegal, Cameroon, Morocco and Madagascar. France, Belgium, Switzerland and Canada together are 21%. This sample covers 9 countries of birth, led by Senegal and Tunisia. 17 of the 20 speakers use French as a…

Publicly accessible cc-by-nc-4.0 n<1K