{"id":"google-gemini-3-5-transcribe","nom":"Gemini 3.5 Transcribe","editeur":"Google","pays":"États-Unis","poids_ouverts":false,"poids_selon":"Artificial Analysis","licence":null,"licence_selon":null,"liens":{"page":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-transcribe","essayer":"https://aistudio.google.com/"},"fiche_auto":true,"page":"https://quelleia.com/voix/modeles/google-gemini-3-5-transcribe/","releve":"2026-10-19","en_bref":"7e sur 51 en transcription en anglais (Artificial Analysis).","presentation":"Modèle de transcription dédié de Google, fondé sur la compréhension audio de Gemini, pour des fichiers jusqu'à une heure, avec une variante en flux continu (Gemini 3.5 Transcribe Live). Il détecte la langue à chaque énoncé, gère les changements de langue, sépare les locuteurs, horodate chaque mot et accepte un vocabulaire personnalisé.","sortie":"2026-08-26","langues":{"annoncees":{"annoncees":"plus de 85 langues, dont le français","nombre":85,"plus":true,"francais":true},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["presentation"],"lu":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-transcribe (« Gemini 3.5 Transcribe is a speech-to-text model based on Gemini's audio understanding capabilities. »), 17/10/2026"},{"appuie":["langues","presentation"],"lu":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-transcribe (« It automatically detects 85+ languages, handles multi-language code-switching, identifies individual speakers, outputs word-level timestamps, and adapts to domain-specific terms through custom vocabulary biasing. »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-transcribe (« Inputs Audio (up to 1 hour) Output Text, Word annotations »), 17/10/2026"},{"appuie":["langues"],"lu":"https://ai.google.dev/gemini-api/docs/models/gemini-3.5-transcribe (« Finnish fi-FI Serbian sr-RS French fr-FR »), 17/10/2026"},{"appuie":["presentation","sortie"],"lu":"https://ai.google.dev/gemini-api/docs/changelog (« August 26, 2026 Gemini 3.5 Transcribe generally available (GA) : Released two dedicated speech-to-text models based on Gemini's audio understanding »), 17/10/2026"}],"classements":[{"id":"transcription/artificial-analysis","nom":"Transcription en anglais, indice AA-WER","rang":7,"sur":51,"valeur":2.6,"unite":"%","chiffres":{"prix_pour_1000_minutes":5,"jeux":{"AA-AgentTalk":2.01,"Earnings22":4.71,"VoxPopuli":1.67},"date_sortie":"2026-08-26","hebergeur":"Google"},"voisins":["stepaudio-3-asr","zoom-scribe-v2-pro"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-text/models/gemini","releve":"2026-10-19"}}],"noms_dans_les_sources":{"artificial_analysis":"Gemini 3.5 Transcribe"}}