Z.AI now supports via a little workaround through Claude Code Desktop via their Coding Plan. You can run Claude Code Desktop with Z.AI's GLM models — glm-4.7, glm-5, glm-5.1, glm-5-turbo, glm-5v-turbo — instead of relying on Anthropic directly.
Here's the complete setup guide.
Download and install the app for your OS. The download link is at code.claude.com/docs/en/desktop.
Open the app → tap the hamburger menu (top left) → Help → Troubleshooting → Enable developer mode. This unlocks the Developer tab in the menu.
Go to Developer → Configure third-Party inference and fill in:
https://api.z.ai/api/anthropicglm-4.7, glm-5.1, glm-5-turbo, glm-5v-turboSave the configuration — the app will restart and validate the settings. A green check means everything is valid.
Open a project or create a new one. In </> Code, you should see your usage tracked against the Z.AI plan. That's how you know it's working.
To switch models per session:
ccd set model glm-5-turbo
Claude Code Desktop uses approximately 100 tokens for context analysis and 30 tokens per response. Monitor your usage on Z.AI's Billing Dashboard to keep costs predictable.
Z.AI's support via a little workaround for Claude Code Desktop means you can run a full AI coding agent with GLM models — no direct Anthropic dependency needed. The Coding Plan covers the API costs, and the setup takes under 5 minutes.
Download Claude Code Desktop, enable developer mode, point it at https://api.z.ai/api/anthropic, and you're running on GLM.