How to Integrate Z.AI Models in Openclaw

Table of contents: [Show]

Want to use Z.AI GLM models with Openclaw? This guide shows you how to connect your Z.AI API key and start using models like zai/glm-5 in minutes.

Quick Start

Z.AI provides access to GLM models through a simple API. Once configured, you can use Z.AI models alongside other providers like GLM Models Openclaw or LiteLLM Models Openclaw.

Step 1: Get Your Z.AI API Key

First, generate an API key from the Z.AI console. Openclaw uses this key under the environment variable ZAI_API_KEY.

Step 2: Configure Openclaw

You have two options for setup:

Option A: Interactive Setup (Recommended)

Openclaw onboard --auth-choice zai-api-key

Option B: Non-Interactive Setup (for scripting)

Openclaw onboard --zai-api-key "$ZAI_API_KEY"

Or with full non-interactive mode:

Openclaw onboard --non-interactive \
--auth-choice zai-coding-global \
--zai-api-key "$ZAI_API_KEY"

Option C: Manual Configuration

Edit your ~/.Openclaw/Openclaw.json file directly:

{
  "env": { "ZAI_API_KEY": "sk-..." },
  "agents": { 
    "defaults": { 
      "model": { "primary": "zai/glm-5" } 
    } 
  }
}

Testing Your Setup

Once configured, test your connection:

Openclaw models list | grep zai

You should see Z.AI models available. Try a quick prompt:

Openclaw ask --model zai/glm-5 "Hello, can you confirm you're working?"

Tip: Z.AI models support the same features as other OpenAI-compatible providers. You can use them for chat, code generation, and more.

Troubleshooting

Authentication errors: Verify your ZAI_API_KEY is correct and exported in your environment.
Model not found: Ensure the Gateway has been restarted after configuration changes with openclaw gateway restart.
Rate limiting: Check your Z.AI console for usage limits and quotas.

Note: If you're new to Openclaw model integration, you might also want to read about Elevated Mode Openclaw for advanced permissions when configuring providers.

Best Practices

Store your API key in environment variables, not hardcoded in config files
Use ~/.Openclaw/.env for sensitive credentials if running as a daemon
Set a fallback model in case Z.AI is temporarily unavailable
Monitor usage through the Z.AI console to track costs

That's it! You're now ready to use Z.AI GLM models with Openclaw.