CloudToLocalLLM 🦞

A local-first companion and desktop capability layer for your AI agent. Private, secure, yours.

Your AI runs on your hardware. The app gives it voice, vision, desktop control, and a secure channel across all your devices — no data leaves your network unless you decide otherwise.

---

Try It Now

🌐 cloudtolocalllm.online — Web app, works in any browser. Log in, that's it.

🐧 Linux — curl -fsSL https://cloudtolocalllm.online/install.sh | bash (coming soon)

📱 Android — APK builds from CI. Grab the latest from Releases.

You need an agent runtime. Hermes is the primary test path. OpenClaw and compatible agent gateways also work.

---

What Sets It Apart

Layer	What it gives you
Secure Channel	Direct line to your agent, synced across devices
Voice + Avatar	Sidecar companion with personality, memory, natural conversation
Desktop Control	Permissioned access to apps, windows, keyboard, files
Vision	Screen awareness, OCR, camera — explicit per-action consent
Device Mesh	Tailscale-first private network across all your machines
Runtime Manager	Manage agents, skills, sessions, tools, diagnostics

---

Architecture

┌────────────────────────────────────────────────────────┐
│                  CloudToLocalLLM App                     │
│  ┌──────────┐  ┌──────────┐  ┌────────────────────┐    │
│  │  Agent   │  │  Avatar  │  │  Desktop Control   │    │
│  │  Channel │  │  Voice   │  │  Vision            │    │
│  └────┬─────┘  └────┬─────┘  └───────┬────────────┘    │
│       └──────────────┼────────────────┘                  │
│                      ▼                                   │
│           ┌──────────────────────┐                       │
│           │   Agent Adapter     │                        │
│           │  Hermes / OpenClaw  │                        │
│           └──────────┬──────────┘                        │
└──────────────────────┼───────────────────────────────────┘
                       │
              ┌────────┴────────┐
              │                 │
         ┌────▼────┐     ┌─────▼──────┐
         │  Local  │     │  Model     │
         │  Agent  │     │  Providers │
         │  Runtime│     │  (Ollama,  │
         │         │     │  LM Studio)│
         └─────────┘     └────────────┘

Technical deep-dive → System Architecture

---

Platforms

Platform	Status
🐧 Linux	✅ AppImage + auto-update daemon
🌐 Web	cloudtolocalllm.online
📱 Android	✅ APK builds from CI
🪟 Windows	🚧 Installer in development
🍎 macOS	📋 Planned

---

Development

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

# Run

flutter run -d linux   # Desktop
flutter run -d chrome  # Web

# Build

flutter build linux --release
flutter build web --release
flutter build apk --release --split-per-abi

Backend Services

cd services/api-backend && npm install && npm run dev
cd services/streaming-proxy && npm install && npm run dev

Full developer guide → docs/development/BUILD_SCRIPTS.md

---

Documentation

📖 docs.cloudtolocalllm.online

Guide	What's in it
User Guide	Features and usage
Setup Guide	Step-by-step installation
Troubleshooting	Common issues
System Architecture	Technical deep-dive
Deployment Guide	Production setup
Security Guide

---

License

MIT — see LICENSE.