{"id":"meta-seamless-m4t-v2-large","nom":"SeamlessM4T-Large v2","editeur":"Meta","pays":"États-Unis","poids_ouverts":true,"poids_selon":null,"licence":"CC BY-NC 4.0","licence_selon":null,"liens":{"page":"https://huggingface.co/facebook/seamless-m4t-v2-large"},"fiche_auto":false,"page":"https://quelleia.com/voix/modeles/meta-seamless-m4t-v2-large/","releve":"2026-10-19","en_bref":"10e sur 16 en prononciation du français.","presentation":"Modèle ouvert de Meta qui regroupe traduction parole-parole, parole-texte, texte-parole et texte-texte, ainsi que la reconnaissance vocale. Sa version v2 adopte l'architecture UnitY2, qui améliore qualité et vitesse de la génération de parole.","sortie":"2023-11-30","taille":"2,3 milliards de paramètres","langues":{"annoncees":{"annoncees":"101 langues en entrée vocale, 96 en texte, 35 en sortie vocale, dont le français","nombre":35,"francais":true},"prises_en_charge_selon_open_tts":96,"evaluees":["fr"]},"preuves":[{"appuie":["presentation"],"lu":"https://huggingface.co/facebook/seamless-m4t-v2-large/raw/main/README.md (« **SeamlessM4T** is our foundational all-in-one **M**assively **M**ultilingual and **M**ultimodal **M**achine **T**ranslation model delivering high-quality translation for speech and text in nearly 100 languages. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/facebook/seamless-m4t-v2-large/raw/main/README.md (« We are releasing SeamlessM4T v2, an updated version with our novel *UnitY2* architecture. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/facebook/seamless-m4t-v2-large/raw/main/README.md (« - 🎤 101 languages for speech input. - 💬 96 Languages for text input/output. - 🔊 35 languages for speech output. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/facebook/seamless-m4t-v2-large/raw/main/README.md (« | fra | French | Latn | Sp, Tx | Sp, Tx | »), 17/10/2026"},{"appuie":["taille"],"lu":"https://huggingface.co/facebook/seamless-m4t-v2-large/raw/main/README.md (« | [SeamlessM4T-Large v2](https://huggingface.co/facebook/seamless-m4t-v2-large) | 2.3B | »), 17/10/2026"},{"appuie":["sortie"],"lu":"https://ai.meta.com/blog/seamless-communication/ (« Introducing a suite of AI language translation models that preserve expression and improve streaming November 30, 2023 »), 17/10/2026"},{"appuie":["sortie"],"lu":"https://ai.meta.com/blog/seamless-communication/ (« All of the models are built on SeamlessM4T v2, the latest version of the foundational model we released in August. »), 17/10/2026"}],"classements":[{"id":"synthese/prononciation","nom":"Prononciation du français","rang":10,"sur":16,"valeur":7.23,"unite":"%","chiffres":{"rtfx":153.63,"taille_milliards":1.67,"langues":96,"clonage":false},"voisins":["nvidia-magpie-multilingual-357m-feb-2026","pocket-tts"],"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"}}],"noms_dans_les_sources":{"hf_audio":"facebook/seamless-m4t-v2-large"}}