{"acceptedBy":{"actor":"ai","name":"Shipped research agent","on":"2026-10-22"},"aliases":[],"change":{"last4":0,"percent":null,"prior4":0},"description":{"editedBy":{"actor":"ai","name":null,"on":"2026-10-22"},"state":"ai_edited"},"href":"/wiki/task-multilingual-speech/","itemCount":7,"key":"task--multilingual-speech","kind":"task","name":"multilingual speech","neighbours":[{"key":"platform--hugging-face","kind":"platform","name":"Hugging Face","relation":"mentioned_with","weight":7},{"key":"task--voice-cloning","kind":"task","name":"voice cloning","relation":"mentioned_with","weight":4},{"key":"concept--real-time-factor","kind":"concept","name":"real-time factor","relation":"mentioned_with","weight":3},{"key":"licence--apache-2-0","kind":"licence","name":"Apache-2.0","relation":"mentioned_with","weight":3},{"key":"concept--word-error-rate","kind":"concept","name":"word error rate","relation":"mentioned_with","weight":2},{"key":"model--omnivoice","kind":"model","name":"OmniVoice","relation":"mentioned_with","weight":2},{"key":"concept--character-error-rate","kind":"concept","name":"character error rate","relation":"mentioned_with","weight":1},{"key":"dataset--emilia","kind":"dataset","name":"Emilia","relation":"mentioned_with","weight":1},{"key":"licence--boson-higgs-tts-3-research-and-non-commercial-license","kind":"licence","name":"Boson Higgs TTS 3 Research and Non-Commercial License","relation":"mentioned_with","weight":1},{"key":"licence--cc-by-nc","kind":"licence","name":"CC-BY-NC","relation":"mentioned_with","weight":1},{"key":"licence--cc-by-nc-4-0","kind":"licence","name":"CC-BY-NC-4.0","relation":"mentioned_with","weight":1},{"key":"maker--boson-ai","kind":"maker","name":"Boson AI","relation":"mentioned_with","weight":1}],"relatedItems":[{"publishedAt":"2026-07-21T00:00:00Z","reason":"task--multilingual-speech","title":"NVIDIA releases MagpieTTS Multilingual 357M v2607, 12 languages with preset voices","url":"/open-weights/nvidia-magpie-tts-multilingual-357m/"},{"publishedAt":"2026-06-04T00:00:00Z","reason":"task--multilingual-speech","title":"Boson AI releases Higgs TTS 3, a 4B zero-shot voice cloning model for 100+ languages","url":"/open-weights/boson-ai-higgs-tts-3/"},{"publishedAt":"2026-03-30T00:00:00Z","reason":"task--multilingual-speech","title":"k2-fsa releases OmniVoice, a 0.6B zero-shot TTS covering 600+ languages","url":"/open-weights/k2-fsa-omnivoice/"},{"publishedAt":"2026-03-26T00:00:00Z","reason":"task--multilingual-speech","title":"Mistral releases Voxtral TTS, a 4B open-weights speech model, non-commercial","url":"/open-weights/mistral-voxtral-tts/"},{"publishedAt":"2026-03-09T00:00:00Z","reason":"task--multilingual-speech","title":"Fish Audio releases S2 Pro, a 5B TTS model for 80+ languages, research licence","url":"/open-weights/fish-audio-s2-pro/"},{"publishedAt":"2026-01-22T00:00:00Z","reason":"task--multilingual-speech","title":"Qwen releases Qwen3-TTS, Apache 2.0 voice cloning in 10 languages","url":"/open-weights/alibaba-qwen3-tts/"},{"publishedAt":"2025-12-11T00:00:00Z","reason":"task--multilingual-speech","title":"FunAudioLLM releases Fun-CosyVoice3-0.5B-2512, an Apache 2.0 voice cloning TTS","url":"/open-weights/funaudio-fun-cosyvoice3-0-5b-2512/"}],"schemaVersion":1,"series26":[0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0],"series8":[0,0,0,0,0,0,0,0],"weeksOfHistory":46}