GLM Coding Plan: glm-5v-turbo users.

[boreal mulch]

Hey there Claw team, I got selected as part of the 5v-turbo users for the Coding Plan. I'm sure the cohort isn't huge, but mostly composed of power users. I've been rocking it on Claude Code / Open Code, it's cheap multimodality and works amazing. I've yet to see any harness implementing the endpoint for coding plan users, rather leaving the z.ai's pay per token API. I'm sure that if this is implemented on an update before other harnesses, you'll get traction from the GLM Coding Plan power users that were selected for 5v usage. That could beget a downstream effect of having Openclaw orchestrated projects with 5v and as you know, money talks -given that 5v gives this cohort multimodality without incurring into more expensive models.

[paper niche]

yep, this is a solid call, and the good news is OpenClaw already has explicit Z.AI Coding Plan support on current builds/docs.

For Coding Plan users, the relevant onboarding choices are:

openclaw onboard --auth-choice zai-coding-global
# or
openclaw onboard --auth-choice zai-coding-cn

Then you can set your default model, for example:

openclaw config set agents.defaults.model.primary "zai/glm-5v-turbo"
openclaw models list --provider zai

Docs: https://docs.openclaw.ai/providers/zai

There’s also an auto-detect zai-api-key path, but the zai-coding-* options are the right fit if you want to force the Coding Plan surface.

If you try it and it’s not there on your install, paste openclaw --version, that’ll tell us fast if you’re just on a build from before it landed.

[boreal mulch]

This is the tits, my clanker couldn't do it before the latest update, glad you have it! Now off to open-source a reverse engineered and improved y-comb startup with 5v.