Indic TTS Unified v1 A large scale, unified collection of speech data for text to speech (TTS) and speech research. This dataset consolidates 17 distinct source datasets into a single, schema normalized resource covering Indian / South Asian languages, plus major European, African, MENA, and Central Asian languages, with over 13.7 million utterances and 26,000+ hours of audio. All audio is resampled to 24 kHz mono . Every row follows an identical schema regardless of source, enabling seamless multi dataset training without per source preprocessing. Dataset Summary Statistic Value Total utterances 13,777,541 Total audio duration 26,300+ hours Languages covered 58+ Audio format 24 kHz, mono, float32 Configs (subsets) 26 Subsets Config Rows Hours Speakers Languages Source Dataset : : : : : orpheus distill neucodec 400 ~2 1 BarryFutureman/orpheus distill neucodec maya distill neucodec 14,000 33.6 12,801 1 BarryFutureman/maya distill data neucodec emodb neucodec 22,043 40.3 5 1 BarryFutureman/EmoDB neucodec expresso neucodec 11,599 10.9 4 1 BarryFutureman/expresso neucodec nonverbal tts 6,222 17.6 2,296 1 deepvk/NonverbalTTS elise 1,194 2.6 1 1 MrDragonFox/Elise elise hindi 1,147 2.4 1…
Runs entirely in your browser via DuckDB-Wasm — this dataset's real data file is loaded once, then queried locally. Nothing is sent to a server.
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy