[{"data":1,"prerenderedAt":173},["ShallowReactive",2],{"repo-phantivia-cortico-world-gpt-sovits-for-coopanion-en":3,"i18n-phantivia-cortico-world-gpt-sovits-for-coopanion-en":170},{"mirrorPool":4,"detail":13,"relatedProjects":49},[5,9],{"key":6,"label":7,"base":8,"type":6},"github","mirror.github","https:\u002F\u002Fgithub.com\u002F",{"key":10,"label":11,"base":12,"type":10},"auto-proxy","mirror.autoProxy","https:\u002F\u002Fgithub.xxlab.tech\u002F",{"owner":14,"repo":15,"repoId":16,"repoUrl":17,"starCount":18,"language":19,"description":20,"languages":21,"screenshots":25,"homepage":26,"pushedAt":27,"ownerAvatar":28,"ownerName":14,"ownerBio":29,"lastSyncTime":30,"lastReleaseTime":31,"tags":32,"license":33,"category":34,"subgroup":35,"releases":36},"Phantivia","cortico-world-gpt-sovits-for-coopanion",1423986824,"https:\u002F\u002Fgithub.com\u002FPhantivia\u002Fcortico-world-gpt-sovits-for-coopanion",1,"TypeScript","Cortico World for Coopanion: the pet's lines spoken with GPT-SoVITS (V5 on NVIDIA GPUs, v2ProPlus on CPU)",{"TypeScript":22,"Shell":23,"JavaScript":24,"CSS":18},93.1,4.5,1.4,[],"","2026-11-04","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F52902663?v=4","NLPer","2026-11-10T18:09:22.676736+00:00","2026-10-18T04:37:06Z",[],"MIT","ai","AI语音工具",[37],{"tag":38,"publishAt":31,"semVer":39,"prerelease":41,"assets":42},"runtime-e91506f-1",[40,40,40],0,false,[43],{"fileName":44,"sizeMb":45,"sha256":46,"downloadCount":47,"originUrl":48},"gpt-sovits-e91506f-1-win-x64-cpu.zip",988.3,"9ce348074e941388544b4db072ee0b9cfc50c9bd420e61f13ee9364fe90152e1",268,"https:\u002F\u002Fgithub.com\u002FPhantivia\u002Fcortico-world-gpt-sovits-for-coopanion\u002Freleases\u002Fdownload\u002Fruntime-e91506f-1\u002Fgpt-sovits-e91506f-1-win-x64-cpu.zip",[50,68,104,134,157],{"owner":51,"repo":52,"repoId":53,"category":34,"subgroup":35,"language":54,"tags":55,"starCount":60,"ownerUid":61,"latestTag":62,"platforms":63,"lastReleaseTime":65,"downloadCount":66,"description":67},"pbakaus","impeccable",1097346685,"Rust",[56,34,57,58,59],"agent-skills","anti-slop","design","slop",109440,43004,"skill-v4.9.0",[64],"other","2026-11-13T06:18:33Z",5463,"A design guidance tool for AI coding agents with 1 skill, 23 commands, 59 deterministic detector rules, and live browser iteration to improve AI-generated frontend design.",{"owner":69,"repo":70,"repoId":71,"category":34,"subgroup":35,"language":72,"tags":73,"starCount":92,"ownerUid":93,"latestTag":94,"platforms":95,"lastReleaseTime":101,"downloadCount":102,"description":103},"debpalash","VoiceStudio",1206390571,"Python",[34,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91],"audiobook","cuda","dubbing","elevenlabs-alternative","huggingface","local-first","mlx","omnivoice-studio","speech-to-text","text-to-speech","transcription","translate","tts","voice-ai","voice-cloning","voice-generation","voicestudio","workflow",109017,4178343,"v0.5.8",[96,97,98,99,100],"windows","mac","linux","android","ios","2026-11-12T08:53:29Z",10774,"A fully-local open-source AI voice studio offering voice cloning, voice design, video dubbing, dictation, transcription, and audiobook creation in 646 languages — an open-source alternative to ElevenLabs.",{"owner":105,"repo":106,"repoId":107,"category":34,"subgroup":35,"language":72,"tags":108,"starCount":127,"ownerUid":128,"latestTag":129,"platforms":130,"lastReleaseTime":131,"downloadCount":132,"description":133},"modelscope","FunASR",569959091,[109,110,111,112,113,114,115,116,117,118,119,120,121,122,82,123,84,124,125,126],"asr","audio","chinese","emotion-recognition","funasr","mcp-server","multilingual-asr","openai-compatible-api","paraformer","punctuation","pytorch","real-time-asr","speaker-diarization","speech-recognition","streaming-asr","vllm","voice-activity-detection","whisper-alternative",21017,109945100,"runtime-torch-v0.1.7",[64],"2026-11-13T09:50:06Z",11,"FunASR is an industrial-grade open-source speech recognition toolkit (Python) covering offline and streaming ASR, VAD, punctuation restoration, and speaker diarization, with OpenAI-compatible serving and MCP integration.",{"owner":135,"repo":136,"repoId":137,"category":34,"subgroup":35,"language":19,"tags":138,"starCount":150,"ownerUid":151,"latestTag":152,"platforms":153,"lastReleaseTime":154,"downloadCount":155,"description":156},"leemysw","yovoice",1371687666,[139,110,140,141,142,143,144,145,146,147,83,86,148,88,149,96],"ai-music","desktop-app","local-ai","macos","mcp","midi","music-generation","offline-ai","skill","voice","voice-cloning-app",2166,40636507,"v0.7.0",[96,97,98],"2026-11-11T12:55:18Z",181,"yovoice is an open-source local-first AI audio and music studio for macOS and Windows, generating expressive speech, cloned voices, lyric-complete songs and multi-track scores fully offline with zero cloud API fees.",{"owner":158,"repo":159,"repoId":160,"category":34,"subgroup":35,"language":161,"tags":162,"starCount":163,"ownerUid":164,"latestTag":165,"platforms":166,"lastReleaseTime":167,"downloadCount":168,"description":169},"Gary23333","opentts",1356280885,"Swift",[],2,120565299,"v5.0.3",[97],"2026-09-03T18:33:50Z",6,"ForgedEcho is a native macOS voice cloning and AI speech production workbench that wraps several open-source speech models in a GUI, supporting voice cloning, emotion control, subtitle dubbing, audiobook generation and local LoRA fine-tuning, all running offline.",{"description":171,"readmeHtml":172},"Want your Coopanion desktop pet to speak? GPTSovits Voice calls local GPT-SoVITS streaming synthesis before each pet_say to speak replies aloud, with GPU and CPU inference tiers plus preset and custom voices, configured under worlds.gpt-sovits.","\u003Ch2 id=\"project-overview\">Project Overview\u003C\u002Fh2>\u003Cp>This extension by Phantivia is the voice World of the Coopanion extension system: before each \u003Ccode>pet_say\u003C\u002Fcode>, Coo calls \u003Ccode>sovits_speak\u003C\u002Fcode> in the same reply, and the extension streams the sentence through local GPT-SoVITS to the sound card. \u003Ccode>sovits_speak\u003C\u002Fcode> returns only after the first audio block hits the card, so the following \u003Ccode>pet_say\u003C\u002Fcode> bubble appears in sync with the voice.\u003C\u002Fp>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Item\u003C\u002Fth>\n\u003Cth>Value\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>Host\u003C\u002Ftd>\n\u003Ctd>Coopanion (desktop pet)\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Engine\u003C\u002Ftd>\n\u003Ctd>GPT-SoVITS api_v2\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Tiers\u003C\u002Ftd>\n\u003Ctd>GPU with V5 turbo, CPU with v2ProPlus, auto by default\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Config\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>worlds.gpt-sovits\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Cblockquote>\n\u003Cp>The pet speaks, and the voice is synthesized locally in real time.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2 id=\"key-features\">Key Features\u003C\u002Fh2>\u003Cul>\n\u003Cli>\u003Cstrong>Two tiers\u003C\u002Fstrong>: with an NVIDIA GPU it uses V5 turbo (first packet 0.5–1.0 s, 3.5 GB VRAM measured); otherwise v2ProPlus on CPU (0.8–1.1 s, 5.8 GB RAM)\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Preset voices\u003C\u002Fstrong>: female and male bundled (generated by MOSS-VoiceGenerator, parameters in \u003Ccode>assets\u002Fvoices\u002Fprompts.json\u003C\u002Fcode>), or upload a 3–10 s custom reference clip\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Three backends\u003C\u002Fstrong>: connect an existing api_v2 via \u003Ccode>serverUrl\u003C\u002Fcode>, point \u003Ccode>runtimeDir\u003C\u002Fcode> at your own GPT-SoVITS folder, or use the panel's managed install (standalone CPython 3.12 and weights, SHA-256 verified)\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Tools and events\u003C\u002Fstrong>: \u003Ccode>sovits_speak(text)\u003C\u002Fcode> queues synthesis, \u003Ccode>sovits_hush()\u003C\u002Fcode> stops and clears, and the \u003Ccode>gpt-sovits.state\u003C\u002Fcode> event broadcasts availability\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Tunable parameters\u003C\u002Fstrong>: speed, volume (0–1), output device name, and prompt language (default \u003Ccode>zh\u003C\u002Fcode>)\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"install-and-enable\">Install and Enable\u003C\u002Fh2>\u003Col>\n\u003Cli>Open the GPTSovits Voice page in Coopanion's advanced mode\u003C\u002Fli>\n\u003Cli>Keep \u003Ccode>tier: auto\u003C\u002Fcode> — a visible \u003Ccode>nvidia-smi\u003C\u002Fcode> GPU picks the GPU tier; first launch compiles CUDA kernels (74 s measured, then about 11 s)\u003C\u002Fli>\n\u003Cli>Warm up with one sentence each in Chinese, English, and Japanese; the service is ready after warmup\u003C\u002Fli>\n\u003Cli>Without the extension or an un-called \u003Ccode>sovits_speak\u003C\u002Fcode>, \u003Ccode>pet_say\u003C\u002Fcode> stays silent and the pet works as usual\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Ch2 id=\"permissions-and-scope\">Permissions and Scope\u003C\u002Fh2>\u003Cul>\n\u003Cli>\u003Cstrong>Local service only\u003C\u002Fstrong>: the backend runs \u003Ccode>runtime\u002Fpython.exe -I -X utf8 -u api_v2.py\u003C\u002Fcode> listening on a free 127.0.0.1 port\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Disk footprint\u003C\u002Fstrong>: managed installs place runtimes and weights under \u003Ccode>runtimes\u002Fgpt-sovits\u002F<version>\u002F<tier>\u002F\u003C\u002Fcode> and \u003Ccode>models\u002Fgpt-sovits\u002Fpretrained_models\u002F\u003C\u002Fcode>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Volume compensation\u003C\u002Fstrong>: v2ProPlus output is quiet (about −27 dBFS peak), so playback multiplies by 8 with soft limiting\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"notes\">Notes\u003C\u002Fh2>\u003Cul>\n\u003Cli>\u003Cstrong>GPU model hates CPU\u003C\u002Fstrong>: V5 turbo on CPU takes 3–4 s for the first packet at a 1.2–1.5 real-time factor, so the CPU tier is pinned to v2ProPlus\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Tier changes need a restart\u003C\u002Fstrong>: \u003Ccode>tier\u003C\u002Fcode> and backend settings take effect after restarting the inference service\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Events over snapshots\u003C\u002Fstrong>: the voice state in environment prompts is only a prefix-rebuild snapshot; live availability follows the \u003Ccode>gpt-sovits.state\u003C\u002Fcode> event\u003C\u002Fli>\n\u003C\u002Ful>\n",1794624044115]