AwNode

Lightweight local gateway for AitherOS. Bridges your local apps to AI backends.

Install

pip install awnode
# container image:
docker pull ghcr.io/aitherium/awnode:latest

# or full stack:
curl -fsSL https://launch.aitherium.com | bash

Quick Start

# Start with auto-detection (finds Genesis, vLLM, Ollama, or cloud)
awnode start

# Run the published container
docker run --rm -p 8090:8090 ghcr.io/aitherium/awnode:latest

# Force specific backend
awnode start --vllm-url http://localhost:8120
awnode start --cloud
awnode start --local

# Check what's connected
awnode status

# Connect to Elysium cloud
awnode connect aither_sk_live_xxxxx

# Deploy via AitherComet
awnode deploy my-service --target docker --strategy rolling

Backend Priority (auto mode)

1. Genesis (localhost:8001) — full AitherOS pipeline (context, memory, agents)
2. vLLM (localhost:8120) — direct GPU inference (OpenAI-compatible)
3. Elysium (cloud) — hosted AitherOS (requires API key)
4. Ollama (localhost:11434) — local CPU/GPU inference
5. Standalone — no LLM, tools-only mode

API

GET  /health          — Health check
GET  /status          — Backend status (what's connected)
POST /chat            — Chat (proxied to best backend)
POST /deploy          — Deploy via AitherComet
POST /connect         — Register with Elysium

Environment Variables

Release Channels

• PyPI: pip install awnode
• Container: ghcr.io/aitherium/awnode
• Source tags: awnode-v*