{"id":"whisper-large-v3","nom":"Whisper Large v3","editeur":"OpenAI","pays":"États-Unis","poids_ouverts":true,"poids_selon":null,"licence":"Apache 2.0","licence_selon":null,"liens":{"poids":"https://huggingface.co/openai/whisper-large-v3","page":"https://huggingface.co/openai/whisper-large-v3"},"fiche_auto":false,"page":"https://quelleia.com/voix/modeles/whisper-large-v3/","releve":"2026-10-19","en_bref":"24e sur 39 en transcription en français ; 30e sur 51 en transcription en anglais (Artificial Analysis) (ex æquo à valeur égale) ; 36e sur 53 en transcription en anglais (Open ASR).","presentation":"Troisième version du grand modèle Whisper d'OpenAI, pour la transcription multilingue et la traduction vers l'anglais, publiée avec ses poids. Elle reprend l'architecture précédente, ajoute le cantonais et a été entraînée sur 1 million d'heures annotées faiblement et 4 millions d'heures pseudo-annotées.","sortie":"2023-11-06","taille":"1,55 milliard de paramètres","langues":{"annoncees":{"annoncees":"99 langues, dont le français","nombre":99,"francais":true},"prises_en_charge_selon_open_tts":null,"evaluees":["fr","en"]},"preuves":[{"appuie":["sortie"],"lu":"https://github.com/openai/whisper/discussions/1762 (« jongwook Nov 6, 2023 Maintainer We're pleased to announce the latest iteration of Whisper, called large-v3 . »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/openai/whisper-large-v3/raw/main/README.md (« The Whisper large-v3 model was trained on 1 million hours of weakly labeled audio and 4 million hours of pseudo-labeled »), 17/10/2026"},{"appuie":["taille"],"lu":"https://huggingface.co/openai/whisper-large-v3/raw/main/README.md (« | large-v3 | 1550 M »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/openai/whisper-large-v3/raw/main/README.md (« language: - en - zh - de - es - ru - ko - fr »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/openai/whisper-large-v3 (« Safetensors 99 languages whisper audio hf-asr-leaderboard Eval Results arxiv: 2212.04356 License: apache-2.0 »), 17/10/2026"}],"classements":[{"id":"transcription/francais","nom":"Transcription en français","rang":24,"sur":39,"valeur":6.24,"unite":"%","chiffres":{"rtfx":339.65,"jeux":{"Common Voice":9.97,"FLEURS":4.84,"MLS":3.9}},"voisins":["soniox-v5-async","whisper-large-v3-turbo"],"source":{"nom":"Open ASR Leaderboard","mention":"Source : Open ASR Leaderboard, Hugging Face","url":"https://huggingface.co/spaces/hf-audio/open_asr_leaderboard","releve":"2026-10-19"}},{"id":"transcription/artificial-analysis","nom":"Transcription en anglais, indice AA-WER","rang":30,"sur":51,"valeur":4.07,"unite":"%","chiffres":{"prix_pour_1000_minutes":1.15,"jeux":{"AA-AgentTalk":3.44,"Earnings22":7,"VoxPopuli":2.41},"date_sortie":"2023-11-06","hebergeur":"fal.ai"},"voisins":["solaria-1","amazon-transcribe"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-text/models/whisper","releve":"2026-10-19"}},{"id":"transcription/open-asr-anglais","nom":"Transcription en anglais, Open ASR","rang":36,"sur":53,"valeur":7.16,"unite":"%","chiffres":{"rtfx":469.79,"jeux":{"AMI-Cleaned":13.63,"Earnings22-Cleaned-AA-chunked":7.99,"Gigaspeech-Cleaned":8.35,"LS Clean":1.56,"LS Other":3.52,"Private (conversational)":15.13,"Private (scripted)":3.61,"SPGISpeech":2.71,"URGENT24":13.73,"Voice Arena Monsoon":3.95,"Voxpopuli-AA-Cleaned":4.53}},"voisins":["fermionresearch-phonon-2","funaudiollm-fun-asr-nano-2512-hf"],"source":{"nom":"Open ASR Leaderboard","mention":"Source : Open ASR Leaderboard, Hugging Face","url":"https://huggingface.co/spaces/hf-audio/open_asr_leaderboard","releve":"2026-10-19"}}],"noms_dans_les_sources":{"hf_audio":"openai/whisper-large-v3","artificial_analysis":"Whisper Large v3"}}