Conversational speech datasets in 60 languages

Every dataset ships with dual-channel audio, time-aligned transcripts, speaker metadata, and documented consent, licensed for commercial ASR, TTS, and voice-AI training. Can’t find your language? We collect to spec.