New LLM Architecture Gallery
I put together a new LLM Architecture Gallery that collects the architecture figures from The Big LLM Architecture Comparison and A Dream of Spring for Open-Weight LLMs in one place.
The goal is to make it easier to quickly browse recent open-weight LLM architectures without jumping back and forth between article sections.
The gallery includes:
- clickable high-resolution architecture figures
- compact fact sheets for each model
- dates,
config.jsonlinks, tech-report links, and from-scratch implementation links where available - short concept explainers for things like GQA, MLA, SWA, QK-Norm, NoPE, Gated DeltaNet, and related ideas
You can find it here:
If you spot an incorrect fact sheet, mislabeled architecture, or broken link, there is also an issue link at the top of the gallery page.
Cite / Share
Short Description
I put together a new LLM Architecture Gallery that collects the architecture figures from my recent comparison articles in one place, together with compact fact sheets and links.
BibTeX
@misc{raschka2026newllmarchitecturegallery,
author = {Raschka, Sebastian},
title = {New LLM Architecture Gallery},
year = {2026},
month = {March},
url = {https://sebastianraschka.com/blog/2026/llm-architecture-gallery.html},
note = {Accessed: 2026-04-26}
}Suggested Share Text
New LLM Architecture Gallery by Sebastian Raschka: https://sebastianraschka.com/blog/2026/llm-architecture-gallery.html
Read Next
The Big LLM Architecture Comparison
Read the main long-form comparison behind many gallery entries.
A Visual Guide to Attention Variants in Modern LLMs
Compare MHA, GQA, MLA, sparse attention, and hybrid attention.
From DeepSeek V3 to V3.2
Dive deeper into sparse attention and recent DeepSeek architecture updates.