{"id":"cofe-ai-flm-audio","nom":"FLM-Audio","editeur":"Cofe AI","pays":"Chine","poids_ouverts":true,"poids_selon":"Artificial Analysis","licence":null,"licence_selon":null,"liens":{"poids":"https://huggingface.co/CofeAI/FLM-Audio","page":"https://huggingface.co/CofeAI/FLM-Audio"},"fiche_auto":false,"page":"https://quelleia.com/voix/modeles/cofe-ai-flm-audio/","releve":"2026-11-09","en_bref":"37e sur 38 en raisonnement à l'écoute (Big Bench Audio) ; 26e sur 31 en tenue de la conversation (Full Duplex Bench).","presentation":"Modèle audio en duplex intégral à poids ouverts, tiré du modèle omnimodal RoboEgo, qui écoute, parle et tient un monologue intérieur en même temps. Il converse en chinois et en anglais et résiste au bruit et aux interruptions.","taille":"7 milliards de paramètres","langues":{"annoncees":{"annoncees":"chinois et anglais","nombre":2,"francais":false,"liste":["zh","en"]},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["presentation","langues"],"lu":"https://huggingface.co/CofeAI/FLM-Audio/raw/main/README.md (« FLM-Audio is a audio-language subversion of [RoboEgo/FLM-Ego](https://arxiv.org/abs/2506.01934v1) -- an omnimodal model with native full duplexity. It simultaneously listens, speaks, and composes internal monologue, delivering low‑latency, duplex conversational responses in both English and Chinese. »), 23/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/CofeAI/FLM-Audio/raw/main/README.md (« FLM‑Audio is robust to noise and user interruptions, prioritizing responsiveness and naturalness. »), 23/10/2026"},{"appuie":["taille"],"lu":"https://huggingface.co/CofeAI/FLM-Audio/raw/main/README.md (« To handle real-time language and audio, FLM-Audio features an LLM-based backbone with 7B parameters »), 23/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/CofeAI/FLM-Audio/raw/main/README.md (« This project is intended for research use only in compliance with applicable laws. For commercial use, please contact us. »), 23/10/2026"}],"classements":[{"id":"voix-a-voix/raisonnement","nom":"Raisonnement à l'écoute, Big Bench Audio","rang":37,"sur":38,"valeur":16,"unite":"%","chiffres":{"hebergeur":"Cofe AI"},"voisins":["nvidia-personaplex-7b-v1","kyutai-moshi"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-speech","releve":"2026-11-09"}},{"id":"voix-a-voix/conversation","nom":"Tenue de la conversation, Full Duplex Bench","rang":26,"sur":31,"valeur":62.02,"unite":"%","chiffres":{"hebergeur":"Cofe AI"},"voisins":["spacexai-grok-voice-fast-1-0","kyutai-moshi"],"source":{"nom":"Artificial Analysis","mention":"Source : Artificial Analysis","url":"https://artificialanalysis.ai/speech-to-speech","releve":"2026-11-09"}}],"noms_dans_les_sources":{"aa_voix_a_voix":"FLM-Audio"}}