🚀 Open Web UI - DXMatrix v1.0.1 – CPU-only Windows Native Build
☕ Support My Work
If you found this project useful, consider buying me a coffee to fuel more DXMatrix drops and open-source magic!
https://buymeacoffee.com/dxmatrix
Your support keeps this project thriving 💜
🚀 A lightweight, CPU-only, Windows 11 native build of Open WebUI, reimagined for DXMatrix-style workflows.
This is a zero-dependency, offline-first fork of Open WebUI designed for:
- 🖥️ Native Windows 11 execution
- 🧩 CPU-only inference (no CUDA/GPU required)
- ❌ No Docker / No WSL / No Linux
- ✅ Ollama or LM Studio local model support
- 🧬 Enhanced UI/UX for fast desktop agent dev
🔗 Download the latest ZIP release here
- Windows 11
- Node.js 18+
- Python 3.10+
- PowerShell 7+
🎯 Run everything in one go!
./start-both-servers.bat
This script will:
✅ Launch the backend with Uvicorn
✅ Start the frontend (SvelteKit / Vite)
✅ Open two terminals so you can monitor both
Once both servers are running, visit:
arduino
Copy
Edit
http://localhost:3000
💡 This is the easiest way to get started — no Docker, no WSL, just native speed and control.
🧰 Manual Setup (Optional)
powershell
Copy
Edit
# Extract the ZIP
cd .\DXMatrix-Win11-1.0.1_CPU
# Create & activate Python virtual env
python -m venv venv
.\venv\Scripts\Activate.ps1
# Install Python backend dependencies
pip install -r backend\requirements.txt
# Start backend
cd backend
uvicorn main:app --host 127.0.0.1 --port 8000 --reload
Then, open a new terminal for frontend:
powershell
Copy
Edit
# From root folder
npm install
npm run dev
Access the UI: http://localhost:3000
🧪 Features (Compared to Upstream)
Feature DXMatrix Edition ✅
Runs natively on Windows ✅
CPU-only (no GPU / CUDA) ✅
No Docker, no WSL ✅
Local Ollama / LM Studio support ✅
Offline-first ✅
Dev-focused tweaks (DXMatrix-ready) ✅
🔗 Credit
This fork is based on the brilliant Open WebUI project by @tjbck. All core functionality is preserved — this build simply retools the experience for Windows-native, no-dependency power users.
🧠 Author
Made with 💜 by @kizashix
🌐 https://matrixlogiclabs.com
📩 ammar@ag38.me
Let's make AI personal, fast, and hacker-friendly again ⚡
python
---