OpenClaw now has a built-in Z.AI provider, so the current GLM setup does not need a hand-written OpenAI-compatible provider block. Authenticate through OpenClaw’s onboarding flow, then set zai/glm-5.3 as the model.
This guide covers authentication, model selection, and usage limits. If OpenClaw is not installed, start with the Windows and WSL install guide.
Choose the Z.AI billing route
Z.AI offers metered API access and a Coding Plan with a points quota. A subscription controls spend, but it is not unlimited. Automated OpenClaw jobs can consume the allowance quickly, so check usage before enabling frequent schedules.
GLM vs DeepSeek for OpenClaw
| GLM (coding plan) | |
| Billing | Flat monthly subscription |
| Best for | Heavy, daily, always-on use |
| Cost certainty | Fixed |
| Caps | Rolling-window usage limits |
If your use is light or unpredictable, DeepSeek is probably cheaper. If you lean on OpenClaw all day, GLM’s plan usually wins.
Step 1: get a Z.AI key
Sign up with Z.AI, choose the Coding Plan or metered API billing, and generate a key. Estimate how often OpenClaw will run before buying a subscription. Light, occasional use can cost less with metered billing.
Step 2: use OpenClaw onboarding
openclaw onboard --auth-choice zai-api-keyThe built-in provider uses the canonical zai ID and stores authentication in OpenClaw’s credential system. This is safer and less fragile than placing a key in a project file.
Step 3: select GLM-5.3
openclaw models list --provider zai
openclaw models set zai/glm-5.3OpenClaw model references use provider/model. Listing the provider first confirms which models your installed version and account can see.
Step 4: test it
Start OpenClaw and hand it a task that uses a capability, not just chat:
Read the files in this folder and tell me which ones changed most recently.A sensible answer means GLM is connected and the assistant can act. If the model talks but can’t reach files, the model is fine and you’re looking at a skill or permission, not GLM.
OpenClaw + GLM checklist
- Z.AI account created; coding plan vs pay-per-token decided
- API key generated and set as an environment variable
- OpenClaw provider set to Z.AI's base URL with a GLM model
- Per-window usage cap checked if you're on the plan
- Tested with a task that touches files
When GLM is the right call
The Coding Plan suits frequent sessions and scheduled jobs when its quota matches your workload. Metered billing is easier to justify for occasional use because you pay only when OpenClaw calls the model.
Verify the active model
Start a new session and inspect its model status. If OpenClaw reports a missing provider key, rerun the Z.AI authentication flow for the affected agent. Authentication stores can be agent-specific, so a key available to one agent may not automatically appear in another.
For the alternatives, compare against DeepSeek, or skip API costs entirely with a local Ollama setup.
