modstart-lib/aigcpanel

AIGCPanel 是一个简单易用的一站式AI数字人系统，支持视频合成、声音合成、声音克隆，简化本地模型管理、一键导入和使用AI模型。

AIGCPanel – A Desktop Hub for AI‑Powered Digital Humans

What it is – AIGCPanel is a cross‑platform desktop application (Windows, macOS, Linux) built with Electron, Vue 3 and TypeScript. It bundles a collection of AI models and tools that let ordinary users create digital‑human videos, synthesize or clone voices, and run a wide range of audio‑/video‑processing tasks. A “Pro/VIP” tier adds visual workflow orchestration and live‑stream interaction features.

Core capabilities

Area	What you can do	Key models / tech
Digital‑human video	Sync lip movements of a person video to any audio (text‑to‑speech or uploaded sound). Supports green‑screen avatars and template management.	MuseTalk, LatentSync, Wav2Lip, Heygem
Voice	• Text‑to‑speech (multiple TTS models, speed/pitch control)\n• Voice cloning from a reference clip\n• Speech‑to‑text with timestamps (ASR)\n• Replace a video’s soundtrack with generated audio	CosyVoice‑300M/‑Instruct, CosyVoice2‑0.5b, FishSpeech, IndexTTS, SparkTTS, GPT‑SoVITS, FunASR
Toolbox (25+ utilities)	Audio: long‑text → audio, subtitle → audio, normalization, replacement, ASR.\nImage: text‑to‑image, image‑to‑image.\nVideo: background addition, smart cut, zoom, annotation, subtitles, speed changes, format conversion, compression, FFmpeg custom commands, batch composition, etc.	Uses the above models plus FFmpeg and custom scripts
Visual workflow (VIP)	Drag‑and‑drop node editor to chain LLM calls, JavaScript snippets, conditionals, file ops, and any toolbox tool. Supports break‑point resume and history tracking.	LogicFlow canvas, built‑in node types
Smart live‑stream (VIP)	Real‑time monitoring of comments on platforms like Douyin, Bilibili, Huya, Douyu, Kuaishou. Auto‑reply via voice/video based on a configurable knowledge base, with profanity filtering.	Integrated with platform APIs
Model management	Install/start/stop local models, view logs, adjust parameters; also connect to remote API services or cloud‑hosted models (no GPU needed).	Supports the listed TTS/ASR/video models
CLI	`aigcpanel` command lets scripts list models, call functions (e.g., text‑to‑image, voice synthesis), manage servers, and run toolbox tools.	Same backend as GUI

Who might use it?

Content creators who want to generate talking‑head videos without recording a person.
Podcasters / educators needing quick voice‑over generation or voice cloning.
Live‑streamers who want automated comment‑driven interactions.
Small teams that need a unified UI for many audio‑video AI utilities without writing code.

How to get started

Download the installer for your OS from aigcpanel.com (exe, dmg, AppImage or deb).
Open the app, go to Model Management and click the one‑click download for the required model pack.
Use the Digital Human tab to upload a portrait video, type or upload audio, and generate a synced output.
Explore the Toolbox for other media tasks, or build a workflow in the Visual Workflow panel (VIP).
Optionally install the aigcpanel CLI from the Settings → Command‑line tool page to script operations.

Technical snapshot

Framework: Electron + Vue 3 + TypeScript
UI: Arco Design + TailwindCSS
Build: Vite
Persistence: better‑sqlite3 (local DB)
Workflow engine: LogicFlow
License: Apache‑2.0 (open source)

Limitations / notes

The “Pro/VIP” features (visual workflow, cloud model service, live‑stream integration) are gated behind a paid tier.
All AI models are run locally unless you opt for the cloud service; a capable GPU is required for many video‑lip‑sync models.
The software includes a legal disclaimer restricting use to lawful activities under Chinese regulations.

In short – AIGCPanel is a ready‑to‑use desktop suite that brings together several open‑source AI models for generating and manipulating digital‑human video, voice, and general media, with optional workflow automation and live‑stream interactivity.