Z.ai released GLM-5.3-Flash, a mid-sized AI model (320B-A18B) that outperforms its predecessor GLM-5.2 and includes multimodal image-reading capabilities. The model is priced at $0.15–0.50 per million tokens—5× cheaper than Gemini-3.7-Flash and 10× cheaper with launch-period 50% discounts. It approaches Claude Opus 4.8 and Gemini-3.7-Flash performance while natively supporting Chinese AI chips, offering strong value for developers building design-heavy applications like web testing.