CloudToLocalLLM

A privacy-first platform to manage and run powerful Large Language Models (LLMs) locally, with an optional cloud relay for seamless remote access.

Key Features • Download & Install • Documentation • Development

---

🚀 Overview

CloudToLocalLLM bridges the gap between secure local AI execution and the convenience of cloud-based management. Designed for privacy-conscious users and businesses, it allows you to run models like Llama 3 and Mistral entirely on your own hardware while offering an optional, secure pathway for remote interaction.

Note: The project is currently in Heavy Development/Early Access. Premium cloud relay features are planned but not yet live.

✨ Key Features

• 🔒 Privacy-First: Run models locally using Ollama. Your data stays on your device by default.
• 💻 Cross-Platform: Native support for Windows and Linux, with a responsive Web interface. macOS support is in progress.
• ⚡ Hybrid Architecture: Seamlessly switch between local models when needed.
• 🔌 Extensible: Integrated with LangChain for advanced AI workflows and vector store support.
• 📊 Monitoring: Optional Sentry integration for error tracking and performance monitoring.
• ☁️ Cloud Infrastructure: Deployed on Azure AKS with provider-agnostic design for future flexibility.

📋 Prerequisites

To use CloudToLocalLLM locally, you only need one thing:

• Ollama: This is the engine that runs the AI models.
  • After installing, pull a model to get started: ollama pull llama3.2

Installed Versions (Verified on Linux)

• Flutter: 3.38.5
• Node.js: 24.12.0
• npm: 11.6.2
• Ollama: 0.13.5
• Git: 2.51.0
• Docker: 28.2.2
• kubectl: v1.35.0

Status: All CLI tools installed and verified. Run flutter doctor, node --version, ollama --version, docker --version, kubectl version --client to confirm.

📥 Download & Install

Windows & Linux

1. Go to the Latest Releases page.
2. Download the installer or executable for your operating system (.exe for Windows, .AppImage or .deb for Linux).
3. Run the installer and launch the application.

Web Version

You can access the latest web deployment directly at: cloudtolocalllm.online

📖 Documentation

Comprehensive documentation is available in the docs/ directory:

• User Guide: Detailed configuration and usage instructions.
• Developer Onboarding Guide: Deep dive into the codebase.
• Troubleshooting: Solutions for common issues.

🛠️ Development

If you are a developer looking to contribute or build from source, follow these steps.

Tech Stack

• Frontend: Flutter (Linux, Windows, Web) - Developed natively in WSL2
• Backend: Node.js (Express.js) - Native Linux runtime
• AI Runtime: Ollama (Windows Host interop via localhost)
• CI/CD: AI-powered orchestration with Kilocode CLI & xAI Grok-Code-Fast-1
• Development: WSL Ubuntu 24.04 (Primary Terminal) & Kiro IDE

Build from Source (WSL Ubuntu 24.04)

Prerequisites: Flutter Linux SDK (3.5+), Node.js (24 LTS), and Git.

1. Clone the Repository:

   git clone https://github.com/CloudToLocalLLM-online/CloudToLocalLLM.git
   cd CloudToLocalLLM

2. Install Dependencies:

   flutter pub get
   (cd services/api-backend && npm install)

3. Run the App:

   flutter run -d linux   # Native Desktop
   # or
   flutter run -d chrome  # Web Interface

For full developer details, see the Developer Onboarding Guide.

AI-Powered CI/CD

CloudToLocalLLM features an innovative unified AI-powered CI/CD system that automatically:

• Analyzes code changes using Kilocode CLI with xAI Grok-Code-Fast-1
• Determines semantic version bumps (patch/minor/major)
• Calculates which platforms need updates (cloud/desktop/mobile)
• Deploys to multiple platforms in a single workflow execution

Key Features:

• Unified Workflow: Single workflow handles analysis, building, and deployment
• Intelligent Platform Detection: AI determines if changes affect web, desktop, or mobile platforms
• Authentication Priority: Auth0 and login changes automatically trigger cloud deployments
• Direct Deployment: No intermediate orchestration or platform branches required
• Comprehensive Status: All deployment status visible in single workflow run
• Manual Overrides: Force deployment or override platform detection when needed

See AI-Powered CI/CD Documentation for detailed information.

🤝 Contributing

We welcome contributions! Please read our Contributing Guidelines and check the Issues tab.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Website • GitHub