GLM-5.3 is now open-weight

736 points · 245 comments on HN · read original →

Points and comments are a snapshot, not live.

Article body wasn't reachable. The HN discussion summary is below.

What commenters are saying

Commenters praise GLM-5.3's quality, comparing it favorably to Opus 4.8 and noting it lacks "Claude talk." Some find it slow through certain providers, especially on Chinese hardware. GLM-5.3-Flash is seen as a capable, cheaper alternative to DeepSeek, though speed and pricing inconsistencies are flagged. Discussion also covers model size trade-offs, tool use costs, and geopolitical differences in compute availability.