模型节点、波形和控制台都在动
V5 locked13910 / 1391110 repos
LIVE OPEN-SOURCE SONIC OS

10 个开源音乐引擎,汇成一个会动的创作中枢

V6 不碰 V5。它把 ACE-Step、YuE、DiffRhythm、MusicGen、Stable Audio、Demucs、RVC 等 GitHub 项目做成可点击、可验收、可接真实运行时的开放引擎矩阵。

模型节点、波形和控制台都在动
LIVE OPEN-SOURCE SONIC OS模型节点、波形和控制台都在动

首屏用生成级 bitmap、Three.js 引擎星图、视频感光带和 html-in-canvas 控制台,信息密度高但不堆文字。

ACE-StepYuEDiffRhythmRVCDemucs
GitHub 引擎矩阵10 repos
1ACE-Step 1.511k stars

Best first local full-song engine candidate for V6 because the repo already exposes API and UI start scripts.

Full songMITadapter-ready
打开 GitHub
2YuE6.3k stars

Use for structured lyrics plus vocal/song generation planning, close to the product goal of prompt-to-complete-song.

Lyrics songApache-2.0wrapper-needed
打开 GitHub
3DiffRhythm2.3k stars

Good V6 fast lane for long-form song experiments and quick ideation once its runtime is wrapped.

Fast songApache-2.0wrapper-needed
打开 GitHub
4LeVo SongGeneration1.7k stars

Use as V6 quality-control reference for song alignment, structure, and multi-preference generation.

Aligned songResearchresearch-lab
打开 GitHub
5SongBloom784 stars

Use as V6 long-arrangement lane for sketch-to-refinement song workflows up to multi-minute outputs.

Long songResearchresearch-lab
打开 GitHub
6AudioCraft / MusicGen23k stars

Already matches the backend MusicGen adapter seam and can serve as a controllable fallback provider.

Text musicMITadapter-ready
打开 GitHub
7Stable Audio 3484 stars

Use as V6 next-gen audio lane for high-fidelity generation research and benchmark comparison.

Hi-fi audioMITwrapper-needed
打开 GitHub
8Demucs10k stars

Core V6 stem engine for vocals, drums, bass, other, and instrumental separation.

Stem splitMITproduction-tool
打开 GitHub
9RVC WebUI36k stars

Core V6 voice lane for private authorized voice training and singing conversion.

Voice modelMITproduction-tool
打开 GitHub
10Amphion9.8k stars

Use as V6 research bench and evaluation toolkit for audio, music, speech, and singing voice conversion.

ToolkitMITresearch-lab
打开 GitHub
V6 ControlACE-Step 1.5
歌曲输出Full song
还没有生成歌曲。
分轨输出0 stems
还没有分轨。
声模输出0 models
还没有声模。
真实生产门禁checking
Local readychecking
Real executionchecking
Current engineadapter-ready
真实接入阻塞model weightsGPU runtimelive audio smoke
ace-stepadapter-ready / not configured
yuewrapper-needed / not configured
diffrhythmwrapper-needed / not configured
levoresearch-lab / not configured
songbloomresearch-lab / not configured
audiocraftadapter-ready / not configured
stable-audio-3wrapper-needed / not configured
demucsproduction-tool / not configured
作品流0 songs
还没有生成歌曲。
任务流0 live

V6 已接入 13910 前端和 13911 后端。

V6 master preview

No jobs yet.
0 exports