coqui-ai/open-speech-corpora

Curated database organized by license type (CC-0, CC-BY, CC-BY-SA, etc.) with standardized metadata including speaker count, duration, and language coverage across 50+ multilingual corpora. Covers specialized datasets for ASR, TTS, and emotional speech synthesis, ranging from single-speaker TTS datasets to large-scale parliamentary and crowdsourced collections. Community-driven with prioritized emphasis on truly open licenses and commercial/research usability.

1,390 stars. No commits in the last 6 months.