GLM-5.3-Flash Released: 320B MIT-Licensed Model Outperforms Predecessor

matei_zaharia · x · 2026-08-26

Zai released GLM-5.3-Flash (Ox Alpha), a 320B-A18B model that is less than half the size of GLM-5.2 yet beats it across every benchmark. The model features native multimodality, a 1M-token context window, and is released under the MIT License. Notably, it runs entirely on Chinese AI chips. Databricks has announced plans to integrate GLM-5.3-Flash for customers quickly.

Related event: Zhipu Open-Sources GLM-5.3-Flash: 320B MoE with 1M Context at a Fraction of the Cost(29 posts)→

Original post →

More from Models

Models channel →