{"id":"stepaudio-3-asr","nom":"StepAudio 3 ASR","editeur":"StepFun","pays":"Chine","poids_ouverts":null,"poids_selon":null,"licence":null,"licence_selon":null,"liens":{"page":"https://platform.stepfun.ai/docs/en/guides/models/stepaudio-3-asr","essayer":"https://audio.stepfun.ai/"},"fiche_auto":true,"page":"https://quelleia.com/voix/modeles/stepaudio-3-asr/","releve":"2026-10-19","en_bref":"6e sur 51 en transcription en anglais (Artificial Analysis).","presentation":"Modèle de reconnaissance vocale de StepFun bâti sur un grand modèle de langage, qui s'appuie sur le contexte pour mieux reconnaître noms propres, termes techniques et homophones. Il vise aussi les audios difficiles comme le chuchotement, le chant ou la musique de fond.","langues":{"annoncees":{"annoncees":"6 langues, dont le français (en préversion)","nombre":6,"francais":true,"liste":["zh","en","ja","ko","fr","es"]},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["presentation"],"lu":"https://platform.stepfun.ai/docs/en/guides/models/stepaudio-3-asr.md (« StepAudio 3 ASR Max is StepFun's largest ASR model to date. Built on a large language model, it fuses high-precision acoustic modeling with that model's context understanding, knowledge, and reasoning »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://platform.stepfun.ai/docs/en/guides/models/stepaudio-3-asr.md (« improve recognition of complex content such as person and place names, drug names, technical terms, and homophones. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://platform.stepfun.ai/docs/en/guides/models/stepaudio-3-asr.md (« supports whispering, very fast speech, slurred liaison, dragged pronunciation, ambient noise, singing, and background music. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://platform.stepfun.ai/docs/en/guides/models/audio.md (« TTS and ASR - Chinese, English, Japanese, Korean, French, Spanish »), 17/10/2026"},{"appuie":["langues"],"lu":"https://platform.stepfun.ai/docs/en/guides/models/audio.md (« Note: All languages except Chinese and English are in preview. More languages are on the way. »), 17/10/2026"}],"classements":[{"id":"transcription/artificial-analysis","nom":"Transcription en anglais, indice AA-WER","rang":6,"sur":51,"valeur":2.5,"unite":"%","chiffres":{"prix_pour_1000_minutes":6.666667,"jeux":{"AA-AgentTalk":2.02,"Earnings22":4.33,"VoxPopuli":1.65},"date_sortie":"2026-09-14","hebergeur":"StepFun"},"voisins":["microsoft-mai-transcribe-1-5","google-gemini-3-5-transcribe"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-text/models/stepfun","releve":"2026-10-19"}}],"noms_dans_les_sources":{"artificial_analysis":"StepAudio 3 ASR"}}