Conversational Speech Generation
Dataset Size
1 million hours
Dialect and Style Options
Japanese - Tokyo - FormalSpanish - Mexico City - SlangFrench (FR) - Paris - Standard
Flexible Lesson Structure
Model Sizes
Tiny: 1B backbone, 100M decoderSmall: 3B backbone, 250M decoderMedium: 8B backbone, 300M decoder
Multiple Speaker Handling
Multiple Tutor Personalities
Objective Metrics
Word Error RateSpeaker SimilarityHomograph DisambiguationPronunciation Consistency
Partial Multilingual Support Planned
Planned for 20+ languages
Shadowing and Translation Drills
Subjective Metrics
Comparative Mean Opinion Score
Supported Language List
AfrikaansArabicArmenianAzerbaijaniBelarusianBosnianBulgarianCantoneseCatalanCroatianCzechDanishDutchEnglishEstonianFinnishFrenchGalicianGermanGreekGujaratiHebrewHindiHungarianIcelandicIndonesianItalianJapaneseKannadaKazakhKoreanLatvianLithuanianMacedonianMalayMandarin ChineseMaoriMarathiNepaliNorwegianPersianPolishPortugueseRomanianRussianSerbianSlovakSlovenianSpanishSwahiliSwedishTagalogTamilThaiTurkishUkrainianUrduVietnameseWelsh