The Silencio voice network
2
The Silencio voice network
250,000+ hours of off-the-shelf voice data.
2,000,000+ contributors in the network.
180+ countries of contributor origin.
150+ distinct languages.
5,000+ hours of new audio added every day.
Captured opt-in, on real consumer devices, in real environments. Rights-cleared for commercial AI training. Documented chain of custody from contributor to delivery.
This is the largest in-the-wild network for multilingual voice data in operation today. We use it to supply (1) verified scripted audio for TTS and voice-cloning, (2) raw multilingual free speech for foundation-model pretraining, and (3) — building now — human-verified conversational and dialect-labeled audio in the languages where production voice AI still does not work.
Catalog and sample packs: https://lnkd.in/e4xDxCDE