Z.aiLaunchAI ModelsAugust 26, 2026

Z.ai releases open-weight GLM-5.3, beating proprietary rivals on agentic benchmarks

Z.ai released GLM-5.3 as open weights on Hugging Face, achieving 84.5% on the CyberGym vulnerability benchmark and a score of 60 on the Artificial Analysis Intelligence Index, tying top open-weights models. It uses the same base as GLM-5.2 with extended post-training.

How this story unfolded

3 weeks · 15 reports · 34 community posts · 49 of 52 shown

  1. Aug 14
  2. Aug 15
  3. Aug 16
  4. Aug 17
  5. Aug 18
  6. Aug 19
  7. Aug 23
  8. Aug 24
  9. Aug 28
  10. Aug 29
  11. Aug 31
  12. Sep 1
  13. Sep 2

More stories today

Open the live feed