Get EdgeCoder
Install the agent. Join the mesh. Earn credits.
Detected: your OSP2P Mesh Network
Your device joins a global peer-to-peer mesh for distributed AI inference.
Earn Credits
Contribute idle compute and earn EdgeCoder credits automatically.
Private & Local-First
Code and prompts stay on your machine. Inference runs locally when possible.
Multi-Platform
macOS, Windows, Linux, iOS, Docker, and VS Code — one mesh everywhere.
macOS
2
Open the .dmg and drag EdgeCoder to Applications. The app bundles the agent — no separate install needed.
3
Sign in with your EdgeCoder account. The app auto-discovers local models and connects to the mesh.
Servers and GPU hosts: one-line install (Linux, macOS, Windows)
Detects every GPU and runs one agent per card with a model sized to its memory. Same options on all three platforms; the packages above run the same installer.
curl -fsSL https://edgecoder.io/install.sh | sudo bash -s -- --token YOUR_TOKENirm https://edgecoder.io/install.ps1 -OutFile install.ps1; .\install.ps1 -- --token YOUR_TOKENThen
sudo edgecoder-install plan shows the node ids to enrol; pass each token with --token <node-id>=<token>.Windows
2
Run the installer. Allow admin access if prompted.
3
Sign in with your EdgeCoder account to connect to the mesh.
Linux (Debian / Ubuntu)
2
sudo dpkg -i EdgeCoder_amd64.deb3
Sign in with your EdgeCoder account to connect to the mesh.
iOS (iPhone / iPad)
1
Download from App Store
TestFlight available now
2
Sign in with your EdgeCoder account.
3
Go to Settings and paste your token:
YOUR_TOKENVS Code Extension
1
Run in the VS Code command line, or search "EdgeCoder" in the Extensions panel.
ext install edgecoder.edgecoder2
Open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P) → EdgeCoder: Configure.
Docker
1
docker run -d --restart unless-stopped \
--name edgecoder \
-e EDGE_RUNTIME_MODE=worker \
-e AGENT_REGISTRATION_TOKEN=YOUR_TOKEN \
ghcr.io/edgecoder-io/edgecoder:latest