What the world's frontier and research models are actually built on - and the slow drift away from pure attention.
Share of notable models by architecture family, by year. Estimated from the models tracked below - directional, not a census.
Read carefully: pure transformers still dominate the frontier. The story isn't replacement - it's the emergence of hybrids at the edges. Bars reflect tracked notable models, not all models in existence.
Sort any column, or tap a Mechanism cell for a plain-English explanation. Every row carries a confidence level - closed models are inference, not fact.
| Model | Family | Mechanism | Year | Confidence | Weights | Source |
|---|
Architectures being studied and wagered on - the directions most likely to reshape the chart above.
Confirmed means the architecture is stated in an official technical report or paper. Inferred means strong community consensus but not officially disclosed. Unknown means genuinely undisclosed - we don't guess.
Closed models stay honest. GPT, Gemini, Claude and Grok internals aren't fully public, so they're marked inferred rather than asserted as fact - with no fabricated source attached.
Every entry is verified against primary sources before publishing. No source, no publish. Where a lab calls a model "hybrid" to mean dual inference modes (Claude, Command A) rather than a hybrid architecture, it stays classified by its actual mechanism.
Open weights are verified live. Open-weight rows carry a Live badge - their architecture is read straight from each model's official config.json on Hugging Face. The page fetches the latest pull over HTTP and falls back to an embedded snapshot offline. Gated and closed models stay human-tracked against their official reports.