{"id":"kokoro-82m-v1-0","nom":"Kokoro 82M v1.0","editeur":"hexgrad","pays":null,"poids_ouverts":true,"poids_selon":null,"licence":"Apache 2.0","licence_selon":null,"liens":{"page":"https://huggingface.co/hexgrad/Kokoro-82M","essayer":"https://huggingface.co/spaces/hexgrad/Kokoro-TTS"},"fiche_auto":true,"page":"https://quelleia.com/voix/modeles/kokoro-82m-v1-0/","releve":"2026-10-19","en_bref":"3e sur 16 en prononciation du français ; 54e sur 96 en synthèse vocale en anglais (Artificial Analysis), ex æquo avec 12 autres modèles dont les marges d'erreur se recouvrent ; 30e sur 38 en synthèse vocale en anglais (TTS Arena).","presentation":"Modèle de synthèse vocale à poids ouverts de 82 millions de paramètres, fondé sur l'architecture StyleTTS 2 et entraîné sur quelques centaines d'heures d'audio sous licence permissive. Son auteur met en avant sa légèreté, son faible coût d'exécution et sa licence Apache qui autorise l'usage commercial.","sortie":"2025-01-27","taille":"82 millions de paramètres","langues":{"annoncees":{"annoncees":"8 langues, dont le français","nombre":8,"francais":true,"liste":["en","ja","zh","es","fr","hi","it","pt"]},"prises_en_charge_selon_open_tts":8,"evaluees":["fr","en"]},"preuves":[{"appuie":["presentation","taille"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/README.md (« is an open-weight TTS model with 82 million parameters. Despite its lightweight architecture, it delivers comparable quality to larger models while being significantly faster and more cost-efficient. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/README.md (« StyleTTS 2: https://arxiv.org/abs/2306.07691 »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/README.md (« Kokoro was trained exclusively on **permissive/non-copyrighted audio data** and IPA phoneme labels. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/README.md (« This is an Apache-licensed model, and Kokoro has been deployed in numerous projects and commercial APIs. »), 17/10/2026"},{"appuie":["langues","sortie"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/README.md (« | **v1.0** | **2025 Jan 27** | **Few hundred hrs** | [**8 & 54**] »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/hexgrad/Kokoro-82M/raw/main/VOICES.md (« [French](#french): 1F »), 17/10/2026"}],"classements":[{"id":"synthese/prononciation","nom":"Prononciation du français","rang":3,"sur":16,"valeur":5.41,"unite":"%","chiffres":{"rtfx":74.4,"taille_milliards":0.08,"langues":8,"clonage":false},"voisins":["fish-audio-s2-pro","supertonic-3"],"source":{"nom":"Open TTS Leaderboard","mention":"Source : Open TTS Leaderboard, Hugging Face","url":"https://huggingface.co/spaces/hf-audio/open_tts_leaderboard","releve":"2026-10-19"}},{"id":"synthese/artificial-analysis","nom":"Synthèse vocale en anglais, Artificial Analysis","rang":54,"sur":96,"valeur":1064.19,"unite":"Elo","bas":1053.19,"haut":1075.19,"votes":5386,"chiffres":{"prix_par_million_caracteres":0.65,"date_sortie":"2025-01-27"},"voisins":["maya-2-flash","nvidia-magpie-multilingual-357m-feb-2026"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/text-to-speech/model-families/kokoro","releve":"2026-10-19"}},{"id":"synthese/tts-arena","nom":"Synthèse vocale en anglais, TTS Arena","rang":30,"sur":38,"valeur":1473,"unite":"Elo","bas":1450,"haut":1496,"votes":1033,"voisins":["voice-ai-text-to-speech-v1","tontaube-v0"],"source":{"nom":"TTS Arena V2","mention":"Source : TTS Arena V2, TTS-AGI","url":"https://huggingface.co/spaces/TTS-AGI/TTS-Arena-V2","releve":"2026-10-19"}}],"noms_dans_les_sources":{"hf_audio":"hexgrad/Kokoro-82M","artificial_analysis":"Kokoro 82M v1.0","tts_arena":"Kokoro v1.0"}}