Z.AI Unveils GLM-5.3-Flash: 320B Native Multimodal Model with 90% Cost Cut

baseten · x · 2026-08-27

Z.AI has released GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series with 320B total parameters (18B active). It approaches Claude Opus 4.8 in coding and agentic benchmarks, offers 90% lower costs than GLM-5.2, and supports 1M token context with native vision. Baseten has launched the model API on day one.

Related event: Zhipu Open-Sources GLM-5.3-Flash, Matching Claude Opus 4.8 at a Tenth the Cost(42 posts)→

Original post →

More from Models

Models channel →