OpenClaw

The AI that actually does things

OpenClaw brings your AI agents directly into your daily conversations on WhatsApp, Telegram, or Matrix. OpenClaw is a self-hosted solution that lets you interact with your personal assistants from anywhere, keeping infrastructure under your control.

This guide will walk you through the setup process and show you how to combine OpenClaw with Cortecs.

Before you begin: Make sure you have generated your Cortecs API key. If not, check out our QuickStart guide.

1. Install OpenClaw

Run the following command in your terminal to install OpenClaw

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

For additional installation options, refer to the OpenClaw documentation.

2. Complete Onboarding

After the install is done, the onboarding process should start automatically. If this is not the case run the following command

openclaw onboard --install-daemon

During onboarding configure the system to your liking. Answer the following options when prompted:

1. Model/auth provider: Skip for now

2. Filter models by provider: All providers

3. Default model: Keep current

3. Add Cortecs Models

In your home directory (or your specified working directory) the folder .openclaw should have been created after onboarding. Within it is a file named openclaw.json that contains your OpenClaw configuration. In this file, you can replace the "agents" section with your model configuration, as shown in the gpt-oss-120b example below.

"models": {
    "mode": "merge",
    "providers": {
      "cortecs": {
        "baseUrl": "https://api.cortecs.ai/v1",
        "apiKey": "YOUR_CORTECS_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "gpt-oss-120b",
            "name": "Gpt Oss 120b (by Cortecs)"
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "cortecs/gpt-oss-120b"
      },
      "models": {
        "cortecs/gpt-oss-120b": {}
      }
   }
}

After saving the new config your gateway should update automatically. Now, if you send a message through your preferred application, you will interact with your specified model. For a quick test use the following command

openclaw tui

This opens a terminal-based chatting interface to quickly send messages to your model.

4. Adding Multiple Models (Optional)

You can configure multiple models using a single API key. To find available models, check our our catalog. Models with the tag OpenClaw are recommended for this use case. Adapt the config again using the JSON below.

"models": {
    "mode": "merge",
    "providers": {
      "cortecs": {
        "baseUrl": "https://api.cortecs.ai/v1",
        "apiKey": "YOUR_CORTECS_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "gpt-oss-120b",
            "name": "Gpt Oss 120b (by Cortecs)"
          },
          {
           "id": "claude-opus4-6",
           "name": "Claude Opus 4.6 (by Cortecs)"
          },
          {
          "id": "minimax-m2",
          "name": "MiniMax M2 (by Cortecs)"
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "cortecs/gpt-oss-120b"
      },
      "models": {
        "cortecs/gpt-oss-120b": {"alias": "oss" },
        "cortecs/claude-opus4-6": {"alias": "opus"},
        "cortecs/minimax-m2": {"alias": "mini"}
      }
   }
}

The alias parameter within the models is optional but it allows you to quickly switch between models using the following command when using the chat

/model oss
/model opus
/model mini

Alternatively, if you already have an instance of OpenClaw running, you can also ask your agent to change the config for you, as can be seen in the image.