{"acceptedBy":{"actor":"ai","name":null,"on":"2026-10-22"},"aliases":[],"change":{"last4":0,"percent":null,"prior4":0},"description":{"editedBy":null,"state":"none"},"href":"/wiki/task-cloning/","itemCount":11,"key":"task--cloning","kind":"task","name":"Cloning","neighbours":[{"key":"platform--hugging-face","kind":"platform","name":"Hugging Face","relation":"mentioned_with","weight":10},{"key":"licence--apache-2-0","kind":"licence","name":"Apache-2.0","relation":"mentioned_with","weight":8},{"key":"task--voice-cloning","kind":"task","name":"voice cloning","relation":"mentioned_with","weight":8},{"key":"task--multilingual-speech","kind":"task","name":"multilingual speech","relation":"mentioned_with","weight":5},{"key":"concept--real-time-factor","kind":"concept","name":"real-time factor","relation":"mentioned_with","weight":3},{"key":"model--qwen3-tts","kind":"model","name":"Qwen3-TTS","relation":"mentioned_with","weight":3},{"key":"concept--vram","kind":"concept","name":"VRAM","relation":"mentioned_with","weight":2},{"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":"platform--github","kind":"platform","name":"GitHub","relation":"mentioned_with","weight":2},{"key":"task--text-to-speech","kind":"task","name":"text-to-speech","relation":"mentioned_with","weight":2},{"key":"task--voice-design","kind":"task","name":"voice design","relation":"mentioned_with","weight":2}],"relatedItems":[{"publishedAt":"2026-08-25T00:00:00Z","reason":"task--cloning","title":"BreezeBlue releases Breeze-TTS-2, a 3B English and Chinese TTS, non-commercial","url":"/open-weights/breezeblue-breeze-tts-2/"},{"publishedAt":"2026-07-21T00:00:00Z","reason":"task--cloning","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--cloning","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":"2026-04-03T00:00:00Z","reason":"task--cloning","title":"OpenBMB releases VoxCPM2, a 2B Apache 2.0 TTS with voice design and 48 kHz output","url":"/open-weights/openbmb-voxcpm2/"},{"publishedAt":"2026-03-30T00:00:00Z","reason":"task--cloning","title":"k2-fsa releases OmniVoice, a 0.6B zero-shot TTS covering 600+ languages","url":"/open-weights/k2-fsa-omnivoice/"},{"publishedAt":"2026-02-12T00:00:00Z","reason":"task--cloning","title":"NineNineSix releases KaniTTS2-en, a 400M English TTS with voice cloning","url":"/open-weights/nineninesix-kani-tts-2-en/"},{"publishedAt":"2026-02-06T00:00:00Z","reason":"task--cloning","title":"OpenMOSS Team releases MOSS-TTS, an 8B Apache 2.0 voice cloning and long-form TTS","url":"/open-weights/openmoss-moss-tts/"},{"publishedAt":"2026-01-22T00:00:00Z","reason":"task--cloning","title":"Qwen releases Qwen3-TTS, Apache 2.0 voice cloning in 10 languages","url":"/open-weights/alibaba-qwen3-tts/"},{"publishedAt":"2025-12-22T00:00:00Z","reason":"task--cloning","title":"Soprano-80M: an 80M Apache 2.0 TTS that claims 2000x real time on GPU","url":"/open-weights/ekwek-soprano-80m/"},{"publishedAt":"2025-12-11T00:00:00Z","reason":"task--cloning","title":"FunAudioLLM releases Fun-CosyVoice3-0.5B-2512, an Apache 2.0 voice cloning TTS","url":"/open-weights/funaudio-fun-cosyvoice3-0-5b-2512/"},{"publishedAt":"2025-11-12T00:00:00Z","reason":"task--cloning","title":"StepFun releases Step-Audio-EditX, a 3B open-weights model for editing and generating speech","url":"/open-weights/stepfun-step-audio-editx/"}],"schemaVersion":1,"series26":[0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0],"series8":[0,0,0,0,0,0,0,0],"weeksOfHistory":50}