Skip to main content

OpenClaw

Run OpenClaw on your Standard Compute budget.

1. Install OpenClaw and the launcher​

npm install -g openclaw
curl -fsSL https://standardcompute.com/cli/install.sh | sh

The launcher requires Node.js 20 or newer. Open a new terminal after installing so standardcompute is on your PATH.

2. Launch OpenClaw​

standardcompute launch openclaw

Choose an API key and approve the connection in your browser. The OpenClaw terminal UI opens automatically.

How the connection works​

The launcher uses a separate, saved Standard Compute profile with its own sessions and workspace. Your usual OpenClaw setup stays untouched, and existing OpenClaw services keep running separately.

The launcher starts a local gateway and opens the terminal UI. Closing the UI stops that gateway. It does not install a background service, connect messaging channels or start scheduled heartbeats.

Manual configuration​

To add Standard Compute to your own OpenClaw config, back it up first:

cp ~/.openclaw/openclaw.json ~/.openclaw/openclaw.json.bak

Then add the provider and make it the default:

~/.openclaw/openclaw.json
{
  "models": {
    "mode": "merge",
    "providers": {
      "standardcompute": {
        "baseUrl": "https://api.stdcmpt.com/v1",
        "apiKey": "YOUR_API_KEY",
        "api": "openai-responses",
        "models": [
          {
            "id": "standardcompute",
            "name": "Standard Compute",
            "reasoning": false,
            "input": ["text", "image"],
            "contextWindow": 1000000,
            "maxTokens": 8192,
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 }
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": { "primary": "standardcompute/standardcompute" },
      "imageModel": { "primary": "standardcompute/standardcompute" },
      "subagents": { "model": "standardcompute/standardcompute" }
    }
  }
}

"mode": "merge" keeps your other providers. Costs are zero because usage is billed against your plan budget, not per token. To keep a previous model as a fallback, add it to fallbacks, for example "model": { "primary": "standardcompute/standardcompute", "fallbacks": ["openai/gpt-4o"] }.

Validate before restarting:

openclaw config validate