Jie Tang's post marking GLM-5.3's API launch says only 'Artificial Analysis Index = 60,' with pricing held level with the earlier GLM-5.2 model. That number is consistent with how Artificial Analysis later scored the cheaper GLM-5.3 Flash variant at 57, three points behind full GLM-5.3, on its intelligence-per-dollar Pareto frontier. Neither figure was checked directly against Artificial Analysis's own GLM-5.3 model page for this item, so the score is treated as Z.ai's self-reported claim, supported by that internal consistency.
