Story · Zixuan Li (Z.ai)

GLM-5.2: Open Weights, Near-Frontier Intelligence (Zixuan Li (Z.ai))

talk · Story page

Li says the open-weight GLM-5.2 lands between Claude Opus 4.7 and 4.8 on the hardest long-horizon coding and agentic benchmarks, and that its non-thinking mode beats GLM-5.1 with thinking turned on. Z.ai also introduced Z Code, its own coding harness.

In plain words

  • Z.ai introduced artificial intelligence software that it says performs close to leading competitors on tests of lengthy programming tasks.
  • Customers can download its learned settings, the information that shapes how it responds, to run it themselves.
  • Z Code, its new programming tool, works with this software and other leading alternatives.
  • Businesses and governments can train the software further to make it better suited to work such as law or finance.

Appeared in

Subscribe

Get the brief in your inbox

Pick daily, weekly, or both. Nothing is gated either way: every issue is on the site and in the feeds.

  • Weekdays at 8:45am IST, one lead story and 6 to 9 items.
  • Sundays, an argued synthesis rather than a recap.
  • One click to leave, and quiet days say so in the subject line.
How often

Weekdays 8:45am IST + Sundays. Unsubscribe in one click.

You're asking for The Agentic Brief by email at the cadence you picked. You can unsubscribe in one click from any issue, and your address is never sold or shared.