Sebastian Raschka's LLM Architecture Gallery now covers 102 models with a diff comparison tool

rasbt · x · 2026-09-04

Sebastian Raschka's LLM Architecture Gallery is a curated reference now covering 102 text-only LLMs (last updated Sep 3). Each card includes architecture diagrams, implementation links, release dates, attention mechanisms, decoder types, and fact sheets. A built-in diff tool lets you compare any two models side by side (e.g., DeepSeek V3.2 vs GLM-5, Qwen3 Next vs MiniMax M2). Multimodal releases are described via their text backbone. Corrections welcome via issues; a print-ready poster version is available.

Original post →

More from Models

Models channel →