Supercharging Openclaw with DGrid: The Ultimate AI Agent Setup

OpenClaw has rapidly emerged as a leading consumer-grade platform for AI agent orchestration. It excels at connecting your digital life — calendars, emails, and notion workspaces — into a seamless, automated workflow. However, an agent is only as powerful as the model driving it.

While Openclaw handles the tools, DGrid provides the intelligence.

By integrating DGrid, you bypass the heavy hardware requirements of running local models (like Ollama) while avoiding the high latency and costs of legacy providers. In this guide, we will show you how to inject DGrid’s high-performance API (powering 200+ models like OpenAI GPT and Gemini) directly into Openclaw, giving you a lightning-fast, reasoning-capable AI assistant on your desktop.

Prerequisites

Before we begin, ensure you have:

• ​Terminal Access​: Terminal (macOS/Linux) or PowerShell (Windows).
• ​DGrid API Key​: Obtain your key from the DGrid dashboard(official guide ).
  • Note: In the configuration steps below, replace ​ sk-Your-DGridAI-Key​ ​with your actual key.
• ​Node.js 22+ ​: Openclaw’s core functionality relies on Node.js, and it requires Node.js version 22 or higher. To check your current Node.js version, run the following command in your terminal: node -v - If your version is lower than 22, update Node.js to meet the requirement before proceeding.

Step 1: Installing Openclaw

First, we need to install the Openclaw core on your machine. This lightweight installation sets up the environment required to run your agents.

For macOS and Linux: Copy and paste the following command into your terminal:

curl -fsSL https://openclaw.ai/install.sh | bash

For Windows (PowerShell): Open PowerShell as an Administrator and run:

iwr -useb https://openclaw.ai/install.ps1 | iex

Wait for the installation to complete (this may take 1–2 minutes).

Step 2: Injecting the DGrid Configuration

With Openclaw installed, configure it to communicate with DGrid’s servers by editing the openclaw.json configuration file.

1. Open the Configuration File

Use the nano editor (or your preferred text editor) to open the config file:

nano ~/.openclaw/openclaw.json

2. Add DGrid Credentials

Add DGrid to the models section and set it as the default in the agents section.

Notes:

• If the file is empty: Paste the full code below.
• If existing configs exist: Merge the keys (preserve other providers if needed, ensure JSON syntax is valid — no missing commas/brackets).

Copy this updated configuration (optimized for GPT/Gemini):

{
  "models": {
    "providers": {
      "DGrid": {
        "baseUrl": "https://api.dgrid.ai/api/v1",
        "apiKey": "sk-Your-DGridAI-Key",
        "api": "openai-responses",
        "models": [        // List all the models you want to use
          {
            "id": "openai/gpt-3.5-turbo",
            "name": "GPT-3.5 Turbo",
            "contextWindow": 16385,
            "maxTokens": 4096
          },
          {
            "id": "google/gemini-1.5-pro",
            "name": "Gemini 1.5 Pro",
            "contextWindow": 1048576,
            "maxTokens": 20480
          }
        ]
      }
    }
  },
  
  // Tell openclaw which model to use by default (format: provider/model ID)
  "agents": {
    "defaults": {
      "model": {
        "primary": "dgrid/google/gemini-1.5-pro"
      }
    }
  }
}

3. Save Changes

• Press Ctrl + O then Enter to save.
• Press Ctrl + X to exit the editor.

Pro Tip: We have set dgrid/google/gemini-1.5-pro as your primary model in the agents section. This ensures that by default, your Openclaw agents use Gemini’s advanced reasoning capabilities for all tasks.

Step 3: Initialization and Onboarding

Now that the “brain” is configured, we need to wake up the “body.” We will run the Openclaw daemon and use the onboarding wizard to lock in our settings.

Run the following command:

openclaw onboard --install-daemon

The Setup Wizard

The system will prompt you for several configurations. To ensure DGrid is prioritized, select the following options:

1. Onboard mode: Select QuickStart.
2. Config handling: CRITICAL — Select Use existing values.
   • Reasoning: This tells Openclaw to respect the JSON file we just edited.
3. Model/auth Provider: Select Skip for now.
4. Filter models by provider: Type or Select DGrid.
5. Default model: Select Keep Current.
   • Reasoning: This confirms the gemini-1.5-pro default we set in the JSON.

Note: You may be asked to install plugins (Skills). For the first run, we recommend skipping these to ensure the connection is stable before adding complexity.

Step 4: Verification

Your Openclaw instance is now running and powered by DGrid (using Gemini 1.5 Pro by default)!

Verify the Connection

To confirm everything works:

1. Open the Openclaw interface.
2. Send a simple query (e.g., “Hello, confirm you’re Gemini 1.5 Pro”).
3. Check the response — Openclaw will confirm it’s using the configured model.

Support

If you encounter any issues with API keys or model usage, please visit the DGrid dashboard or contact our support team .

Thank you for choosing DGrid. Happy coding!