GLM-5.3-Flash matches Opus 4.8, 45x cheaper

Hesamation · x · 2026-08-26

Evaluations reveal that Zhipu's GLM-5.3 Flash scores 57 on the Artificial Analysis intelligence index, matching Opus 4.8. However, it costs only $0.045 per task, making it 45x cheaper than Opus 4.8 and 15x cheaper than GLM-5.3. It also achieves 3x lower attention compute and 4.4x smaller KV cache at 1M context.

Related event: Z.ai Open-Sources GLM-5.3-Flash: 320B MoE Rivals Claude Opus 4.8 at Fraction of Cost(26 posts)→

Original post →

More from Models

Models channel →