Repolode
Submit
techjarves

techjarves/Portable-Local-Studio

Offline desktop-style local AI studio bundling Stable Diffusion image generation, GGUF chat, Whisper speech-to-text and Kokoro text-to-speech.

View on GitHub
Stars
1,942
Forks
469
Open issues
6
Last commit
3 days ago

What is it?

Portable AI Studio is a self-contained local AI app for Windows, Linux and macOS that runs image generation, LLM chat, speech-to-text and text-to-speech in one browser-based interface. The runtime, including Node.js and the GPU backends, is bundled in the project folder, so setup does not change global system settings. Launching the included windows.bat, linux.sh or mac.sh script opens the UI at localhost:1420.

Image generation uses stable-diffusion.cpp with single-file SD 1.5 and SDXL checkpoints in .safetensors, .ckpt or GGUF formats. Text chat runs GGUF models through a llama.cpp server, and speech recognition uses whisper.cpp models. Text-to-speech relies on the Kokoro-82M model via kokoro-js. A built-in Model Manager can download models from Hugging Face URLs or import local files by drag and drop.

The app auto-detects hardware and selects CUDA for Nvidia, ROCm or Vulkan for AMD and Intel, Metal on Apple Silicon, or OpenVINO for Intel Core Ultra NPUs, with CPU as fallback. Text and image engines are kept mutually exclusive by default to limit memory use.

Key features

  • 01Runs Stable Diffusion image generation, GGUF chat, Whisper transcription and Kokoro speech synthesis offline.
  • 02Portable runtime bundled in the project folder with no global system changes.
  • 03Auto-detects CUDA, ROCm, Vulkan, Metal and OpenVINO backends, with CPU fallback.
  • 04Model Manager downloads from Hugging Face URLs or imports local weights by drag and drop.
  • 05Built-in performance monitor shows CPU, RAM, GPU and VRAM usage.
  • 06Saves generated images with prompt parameters and metadata JSON files.

Use cases

  • Chatting with a GGUF language model without sending prompts to a cloud service.
  • Generating images with local SD 1.5 or SDXL checkpoints on your own GPU.
  • Transcribing voice recordings offline with Whisper models.
  • Running a portable AI setup on a USB drive or a machine you cannot install software on.

Quick start

  1. 1

    Windows: double-click windows.bat (first run downloads a portable Node.js runtime and backends)

  2. 2

    Linux: make the launcher executable

    chmod +x linux.sh
  3. 3

    Linux: launch the studio

    ./linux.sh
  4. 4

    Open http://localhost:1420 in your browser, then add a model to app/models/ or use the Model Manager

Things to know

  • macOS Intel hardware is completely unsupported; the macOS backend targets Apple Silicon (M1 or newer).
  • Prebuilt Linux backends require glibc 2.38+ and GLIBCXX_3.4.32+, so older distributions may fail.
  • Without a supported GPU, image generation runs on CPU threads and is slow.
  • SDXL models need considerably more RAM and VRAM than SD 1.5 models.

In collections

Run LLMs locallyPrivacy & local-first toolsAI video & audio toolsAI image generators & Stable Diffusion tools

Topics

#local-ai#stable-diffusion#llm#whisper#text-to-speech#speech-to-text#offline#gguf#image-generation#text-generation

FAQ

What is Portable-Local-Studio?
Offline desktop-style local AI studio bundling Stable Diffusion image generation, GGUF chat, Whisper speech-to-text and Kokoro text-to-speech. Portable AI Studio is a self-contained local AI app for Windows, Linux and macOS that runs image generation, LLM chat, speech-to-text and text-to-speech in one browser-based interface. The runtime, including Node.js and the GPU backends, is bundled in the project folder, so setup does not change global system settings. Launching the included windows.bat, linux.sh or mac.sh script opens the UI at localhost:1420.
Is Portable-Local-Studio free to use? What license does it have?
Portable-Local-Studio is open source under the MIT license. Check the license terms before using it in commercial products.
Is Portable-Local-Studio still maintained?
Actively maintained. The last commit to Portable-Local-Studio was on Nov 5, 2026.
What can I use Portable-Local-Studio for?
Chatting with a GGUF language model without sending prompts to a cloud service. Generating images with local SD 1.5 or SDXL checkpoints on your own GPU. Transcribing voice recordings offline with Whisper models. Running a portable AI setup on a USB drive or a machine you cannot install software on.
How do I install Portable-Local-Studio?
Windows: double-click windows.bat (first run downloads a portable Node.js runtime and backends) Linux: make the launcher executable: chmod +x linux.sh Linux: launch the studio: ./linux.sh Open http://localhost:1420 in your browser, then add a model to app/models/ or use the Model Manager
Where can I get Portable-Local-Studio?
The source code and installation instructions are on GitHub: https://github.com/techjarves/Portable-Local-Studio

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects