Model providers

Using Z.AI with OpenClaw

3 min read

Browse more in Model providers.

All model providers guides →

AI as a model provider inside OpenClaw. AI API key, pick the right regional or Coding Plan endpoint, and set a default GLM model for your agents.

1` through the `zai` provider.

Prerequisites

  • ✓A Z.AI account with access to GLM models and an API key created in the Z.AI console.
  • ✓An existing OpenClaw installation with the `openclaw` CLI available in your shell.
  • ✓Network access from your OpenClaw environment to the Z.AI Chat Completions API.

Steps

  1. 1

    Decide between auto-detect and explicit regional onboarding

    AI supports both an auto-detected endpoint based on your key and explicit regional or Coding Plan endpoints. Decide upfront whether you want OpenClaw to infer the correct base URL from the key prefix or force a specific Coding Plan or region.

    AI.

  2. 2

    Run onboarding with auto-detected Z.AI endpoint

    AI endpoint from your API key. This is best for most setups because it configures the correct base URL automatically from the key prefix.

    Run this in the environment where your OpenClaw agents execute.

    bash
    openclaw onboard --auth-choice zai-api-key
  3. 3

    Run onboarding with an explicit Z.AI regional or Coding Plan endpoint

    If you need to force a specific Coding Plan or general API surface, use one of the explicit onboarding choices instead of auto-detect. These commands lock OpenClaw to the global or China region and to either the Coding Plan or general API.

    AI account is provisioned.

    bash
    # Coding Plan Global (recommended for Coding Plan users)
openclaw onboard --auth-choice zai-coding-global

# Coding Plan CN (China region)
openclaw onboard --auth-choice zai-coding-cn

# General API
openclaw onboard --auth-choice zai-global

# General API CN (China region)
openclaw onboard --auth-choice zai-cn
  4. 4

    Set a default Z.AI GLM model for your agents

    AI’s GLM models without specifying a model on every call. 1` as the primary model and wires in your `ZAI_API_KEY`.

    defaults`.

    json
    {
  env: { ZAI_API_KEY: "sk-..." },
  agents: { defaults: { model: { primary: "zai/glm-5.1" } } },
}
  5. 5

    Verify Z.AI models are registered in OpenClaw

    AI catalog. Listing models for the `zai` provider helps you catch auth or region issues early.

    Run this from the same environment where you configured the provider.

    bash
    openclaw models list --provider zai
  6. 6

    Optionally disable Z.AI tool-call streaming for specific models

    AI enables `tool_stream` by default for tool-call streaming, which may not fit every integration. You can override this behavior per model by setting `tool_stream: false` under the `zai/<model>` entry.

    Use this when your tools or downstream consumers expect non-streaming tool calls.

    json
    {
  agents: {
    defaults: {
      models: {
        "zai/<model>": {
          params: { tool_stream: false },
        },
      },
    },
  },
}

Configuration

OptionDescriptionExample
ZAI_API_KEYZ.AI API key used by the `zai` provider for Bearer authentication against the Z.AI Chat Completions API.sk-...
agents.defaults.model.primarySets the default primary model for all agents, here pointing to a Z.AI GLM model reference.zai/glm-5.1
agents.defaults.models."zai/<model>".params.tool_streamControls whether tool-call streaming is enabled for a specific Z.AI model; `tool_stream` is enabled by default.false

Troubleshooting

Z.AI models do not appear when running `openclaw models list --provider zai`

This usually means onboarding did not run with the correct auth choice or the `ZAI_API_KEY` is missing from your environment. " }`.

bash
openclaw onboard --auth-choice zai-api-key

Tool calls from Z.AI models stream when your integration expects non-streaming behavior

AI enables `tool_stream` by default for tool-call streaming, which can break consumers that assume a single tool result payload. models` for `"zai/<model>"` and set `params: { tool_stream: false }`.

bash
{
  agents: {
    defaults: {
      models: {
        "zai/<model>": {
          params: { tool_stream: false },
        },
      },
    },
  },
}

Frequently asked questions

Powered by Mem0

Add persistent memory to OpenClaw

Official Mem0 plugin for OpenClaw keeps context across chats and tools. Smaller prompts, lower cost, better continuity for your agents.

Read the official Mem0 guideOpen Mem0

More in Model providers

Model providers

Using Alibaba Model Studio with OpenClaw

Model providers

Using BytePlus with OpenClaw

Model providers

Using fal with OpenClaw

Model providers

Using GLM (Zhipu AI) with OpenClaw

Model providers

Using Hugging Face with OpenClaw

Model providers

Using Inferrs with OpenClaw