LLM Architectures Explained

How language-model architectures differ, model by model.

Every large language model is a stack of the same few parts, chosen differently: how attention shares its keys and values, how positions are encoded, where the norms sit, whether the feed-forward block is one network or many experts, how deep and how wide. This site records those choices for 160 models, each value traced to the model’s own configuration, paper or announcement, and computes what they cost.

models

160

gallery checklist names covered

109/109

configs pinned to a Hugging Face commit

122

closed models, sizes not disclosed unless stated

23

Part of a family of companion sites: the Transformer Decoder Explainer (one forward pass), LLM Inference Explained (serving it), GPU Kernels Explained (how a GPU executes it), Numerics Explained (the number formats it runs in) and Systolic Arrays Explained (the matrix hardware of TPUs). How this site was built, and how to check its data: about.