Your own AI coding cloud.
A real Linux workspace in your browser — opencode, Claude Code, Codex and more, behind one key and one OpenAI-compatible endpoint. Rolling usage windows. No per-token invoices.
Built for agents, priced for humans.
One key, every lane
The whole catalog behind one OpenAI-compatible endpoint. Swap models with one line and keep your agent, editor, and scripts exactly as they are.
Windows, not invoices
Usage runs in rolling daily and monthly windows per plan. The runway chart below is live gateway data — what you see is what you get.
Your key, your cost
Paste your own API key and every CLI uses it through the proxy. No key? The free pool still answers — demo anytime, zero config.
Three steps to your first session
Sign in
Magic link login. No password. Create an account on sign-in.
→ aethercode.app
→ type your email
→ click the link
Create workspace
A real Linux container with all 5 AI coding CLIs pre-installed — boots in seconds.
→ + New workspace
→ wait for "running"
→ open the Terminal
Pick a model
Every CLI talks to OmniRoute. Simulated out of the box, real with one key.
$ opencode run "hello"
$ opencode /model "Kimi K3"
Estimated model runway
LIVE COUNTS FROM THE OMNIROUTE GATEWAY — EVERY CLI REQUEST IS LOGGED. SIMULATED MODE COUNTS TOO.
One key. Any coding agent.
AetherCode is a standard gateway: an OpenAI-compatible endpoint plus an Anthropic-compatible lane. Point any harness at it as a custom provider with your API key (Dashboard → API Keys).
// one line swap keeps your agent exactly as it is { "provider": { "aethercode": { "npm": "@ai-sdk/openai-compatible", "options": { "baseURL": "https://omniroute.aethercode.io/openai", "apiKey": "wrk_..." }, "models": { "claude-haiku-4-5": {}, "gpt-4o-mini": {}, "deepseek-chat": {} } } } }
export ANTHROPIC_BASE_URL="https://omniroute.aethercode.io/anthropic" export ANTHROPIC_API_KEY="wrk_..."
Pick your lane
- 1 workspace
- 30h workspace time / month
- 15 min idle auto-stop
- 2 CLI slots
- 8 hours/day · 240h/month
- 25 GB persistent disk
- One key · every CLI
- AI usage windows · no token bills
- 4 GB RAM
- 50 GB persistent disk
- 720h workspace time / month
- Your own VPS
- 8 GB RAM
- Unlimited usage