techjarves/Portable-Local-Studio

Portable local AI studio for Windows, Linux, and macOS. Zero-setup GUI for Image Generation, GGUF LLMs, Text to Speech & Speech to Text

What it solves

Portable AI Studio provides a zero-configuration, 100% offline environment for running multiple AI modalities on a single machine. It eliminates the need for complex environment setups, cloud subscriptions, and internet connectivity, allowing users to generate images, chat with LLMs, and process speech privately on their own hardware.

How it works

The project bundles several high-performance C++ backends into a self-contained runtime. It uses a web-based GUI to manage these engines, which are mutually exclusive by default to optimize VRAM and RAM usage. It automatically detects hardware to apply the best acceleration method available:

  • Image Generation: Powered by stable-diffusion.cpp supporting CUDA, ROCm, Vulkan, Metal, and OpenVINO.
  • Text Chat: Powered by llama.cpp for GGUF models.
  • Speech-to-Text: Powered by whisper.cpp.
  • Text-to-Speech: Powered by kokoro-js using the Kokoro-82M ONNX model.

Who it’s for

It is designed for users on Windows, Linux, and macOS (Apple Silicon) who want a private, uncensored AI studio without needing to manage Python environments or API keys.

Highlights

  • Zero-Install Portability: Includes its own Node.js runtime and pre-compiled binaries to avoid global system changes.
  • Multi-Modal Integration: Unifies image generation, LLM chat, transcription, and voice synthesis in one interface.
  • Hardware Auto-Detection: Supports a wide range of accelerators including Nvidia (CUDA), AMD (ROCm/Vulkan), Intel (Vulkan/OpenVINO NPU), and Apple (Metal).
  • Integrated Model Manager: Allows users to import weights via Hugging Face URLs or local drag-and-drop.

Related

  • Project
    morettt/my-neuro

    A comprehensive workbench for creating personalized, human-like AI companions with customizable voice, personality, and Live2D visuals.

  • Project
    Devin-AXIS/deepseek-design

    A native visual design system for DeepSeek Harness that allows AI to generate and manually refine editable websites, presentations, and videos.

  • Project
    AutoArk/TinyEngram

    An open research project exploring Engram-based memory injection for LLMs and Stable Diffusion to enable parameter-efficient knowledge updates without catastrophic forgetting.

  • Project
    xandergos/terrain-diffusion

    A learned successor to Perlin noise that uses diffusion models to generate infinite, deterministic, and randomly-accessible real-time terrain and climate data.