GLM pre-training architecture
Z.ai's GLM pre-training architecture and model family. The official about page says GLM was introduced in 2021 and is the company's proprietary pre-training architecture, with later GLM releases built on this line.
Family of models with token-based pricing per 1M tokens (e.g. GLM-5: input $1, cached input $0.2, output $3.2; GLM-4.7-Flash free; image gen GLM-Image $0.015/image). Prices vary by specific GLM variant. Some free tiers/promos.
Pricing varies across GLM variants; price fields are left at zero unless grounded for the general 'GLM' label. Source is the official Z.AI documentation.