How to Configure Model Providers in OpenClaw

I have often found myself staring at a blank configuration file, wondering which API key goes where. It is frustrating to spend more time on setup than on actually building something useful. I prefer a setup that stays out of my way so I can focus on the logic of my project.

OpenClaw makes this part simple by supporting a wide range of providers. You just need to pick one, authenticate, and point the config to your model of choice.

What You’ll Need

• An account with an LLM provider (such as Venice, OpenAI, or Anthropic).
• Authentication credentials for your chosen provider.

Quick Start

You can get your models running in about five minutes. I recommend using Venice AI if you want a privacy-first setup.

1. Authenticate: Run the following command in your terminal to set up your credentials:

   openclaw onboard

2. Set your default model: Open your configuration and define your primary model. Use the format provider/model.

{
  agents: { defaults: { model: { primary: "anthropic/claude-opus-4-6" } } },
}

If you are using Venice, I suggest these models:

• Default: venice/llama-3.3-70b
• Hard tasks: venice/claude-opus-45 (Opus is still the strongest option here).

Troubleshooting

If you are looking for help with specific chat channels like WhatsApp or Telegram rather than model providers, check the Channels documentation. For issues with specific API keys or advanced configurations, refer to the individual provider links in the next section.

If you hit a wall, you can always ask the AI Setup Assistant.

What’s Next

• Channels
• Model providers