Z.ai doubles it's coding plan prices

I can't find an announcement on this, but here's the archived subscription page from a couple days ago: https://web.archive.org/web/20260410092340/https://z.ai/subs...

The Max plan has doubled, and the Lite and Pro plans have more than doubled. No change in usage limits.

Unjustified from the perspective of a consumer. Coding api is incredibly slow, frequently times out, and model quality is nowhere near Claude Opus in actual usage. Usage quotas also fill much sooner than expected given the claimed ratios to Claude plans. No value compared to Codex or Claude. Quite a surprising change.

Anybody running GLM 5.1 locally or within their company? What kind of hardware do you need to achieve performance similar to Claude?

its