{"id":"canary-qwen-2-5b","nom":"Canary Qwen 2.5B","editeur":"NVIDIA","pays":"États-Unis","poids_ouverts":true,"poids_selon":null,"licence":"CC BY 4.0","licence_selon":null,"liens":{"page":"https://huggingface.co/nvidia/canary-qwen-2.5b","essayer":"https://huggingface.co/spaces/nvidia/canary-qwen-2.5b"},"fiche_auto":false,"page":"https://quelleia.com/voix/modeles/canary-qwen-2-5b/","releve":"2026-10-19","en_bref":"34e sur 51 en transcription en anglais (Artificial Analysis) ; 18e sur 53 en transcription en anglais (Open ASR).","presentation":"Modèle de transcription anglaise de NVIDIA qui associe l'encodeur de canary-1b-flash au modèle de langage Qwen3-1.7B, avec ponctuation et majuscules. Il fonctionne en mode transcription ou en mode LLM, où il peut résumer le texte transcrit ou répondre à des questions sur celui-ci.","sortie":"2025-07-17","taille":"2,5 milliards de paramètres","langues":{"annoncees":{"annoncees":"anglais seulement","nombre":1,"francais":false,"liste":["en"]},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["langues","presentation"],"lu":"https://huggingface.co/nvidia/canary-qwen-2.5b/raw/main/README.md (« Canary-Qwen-2.5B is an English speech recognition model »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/nvidia/canary-qwen-2.5b/raw/main/README.md (« The model works in two modes: as a transcription tool (ASR mode) and as an LLM (LLM mode). »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/nvidia/canary-qwen-2.5b/raw/main/README.md (« It is built using two base models: `nvidia/canary-1b-flash` [1,5] and `Qwen/Qwen3-1.7B` [4] »), 17/10/2026"},{"appuie":["taille"],"lu":"https://huggingface.co/nvidia/canary-qwen-2.5b/raw/main/README.md (« With 2.5 billion parameters and running at 418 RTFx »), 17/10/2026"},{"appuie":["sortie"],"lu":"https://huggingface.co/nvidia/canary-qwen-2.5b/raw/main/README.md (« Huggingface 07/17/2025 via https://huggingface.co/nvidia/canary-qwen-2.5b »), 17/10/2026"}],"classements":[{"id":"transcription/artificial-analysis","nom":"Transcription en anglais, indice AA-WER","rang":34,"sur":51,"valeur":4.27,"unite":"%","chiffres":{"prix_pour_1000_minutes":0.74,"jeux":{"AA-AgentTalk":3.21,"Earnings22":6.31,"VoxPopuli":4.34},"date_sortie":"2025-07-17","hebergeur":"Replicate"},"voisins":["modulate-stt-batch-english-vfast","google-chirp-3"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-text/models/nvidia","releve":"2026-10-19"}},{"id":"transcription/open-asr-anglais","nom":"Transcription en anglais, Open ASR","rang":18,"sur":53,"valeur":5.94,"unite":"%","chiffres":{"rtfx":861.56,"jeux":{"AMI-Cleaned":7.91,"Earnings22-Cleaned-AA-chunked":6.41,"Gigaspeech-Cleaned":7.8,"LS Clean":1.23,"LS Other":2.63,"Private (conversational)":13.41,"Private (scripted)":3.47,"SPGISpeech":1.7,"URGENT24":13.02,"Voice Arena Monsoon":3.6,"Voxpopuli-AA-Cleaned":4.14}},"voisins":["edge0-ark-asr-0-6b","gpt-transcribe"],"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":{"artificial_analysis":"Canary Qwen 2.5B","hf_audio":"nvidia/canary-qwen-2.5b"}}