{"id":"step-audio-editx-mar-2026","nom":"Step Audio EditX (Mar 2026)","editeur":"StepFun","pays":"Chine","poids_ouverts":true,"poids_selon":null,"licence":null,"licence_selon":null,"liens":{"page":"https://huggingface.co/stepfun-ai/Step-Audio-EditX","essayer":"https://huggingface.co/spaces/stepfun-ai/Step-Audio-EditX"},"fiche_auto":true,"page":"https://quelleia.com/voix/modeles/step-audio-editx-mar-2026/","releve":"2026-10-19","en_bref":"39e sur 96 en synthèse vocale en anglais (Artificial Analysis), dans un groupe de 10 modèles trop proches pour être départagés.","presentation":"Modèle audio de StepFun spécialisé dans l'édition itérative d'une voix (émotion, style de parole, sons paralinguistiques comme la toux ou les pleurs). Il fait aussi de la synthèse vocale avec clonage de voix sans entraînement préalable (zero-shot).","taille":"3 milliards de paramètres","langues":{"annoncees":{"annoncees":"mandarin, anglais, sichuanais, cantonais, japonais et coréen ; français prévu","francais":false,"liste":["zh","en","ja","ko"]},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["presentation","taille"],"lu":"https://huggingface.co/stepfun-ai/Step-Audio-EditX (« We are open-sourcing Step-Audio-EditX, a powerful 3B-parameter LLM-based Reinforcement Learning audio model specialized in expressive and iterative audio editing. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/stepfun-ai/Step-Audio-EditX (« It excels at editing emotion, speaking style, and paralinguistics, and also features robust zero-shot text-to-speech (TTS) capabilities. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/stepfun-ai/Step-Audio-EditX (« Excellent zero-shot TTS cloning for Mandarin, English, Sichuanese, and Cantonese. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/stepfun-ai/Step-Audio-EditX (« Now supporting Japanese and Korean languages. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/stepfun-ai/Step-Audio-EditX/raw/main/README.md (« - [x] Japanese, Korean - [ ] Arabic, French, Russian, Spanish, etc. »), 17/10/2026"}],"classements":[{"id":"synthese/artificial-analysis","nom":"Synthèse vocale en anglais, Artificial Analysis","rang":39,"sur":96,"valeur":1099.77,"unite":"Elo","bas":1087.77,"haut":1111.77,"votes":2249,"chiffres":{"date_sortie":"2026-02-12"},"voisins":["openai-tts-1-hd","elevenlabs-turbo-v2-5"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/text-to-speech/model-families/stepfun","releve":"2026-10-19"}}],"noms_dans_les_sources":{"artificial_analysis":"Step Audio EditX (Mar 2026)"}}