Now tracking · The machines

What the frontier costs

Where model-building actually is, refreshed from the primary sources rather than from anybody's summary of them.

5 measuresRefreshed MondaysLast refreshed 16 September 2026, 00:23 UTC
4.6×/yr

Frontier training compute

Doubling every 5.4 months. Least-squares fit over 33 frontier models released since 2020, computed from Epoch AI's notable-models dataset.

Epoch AIFetched 16 Sep 2026
5.0e26 FLOP

Largest training run

Grok 4, published 2025-07-09 — the largest training run Epoch AI has been able to document.

Epoch AIFetched 16 Sep 2026
$388M

Cost of a frontier run

Grok 4, published 2025-07-09. Amortised hardware and energy for a single run, in constant 2023 dollars, as estimated by Epoch AI.

Epoch AIFetched 16 Sep 2026
2 of 17

Frontier models with open weights

12% of frontier models released since 2024 ship open weights, per Epoch AI's labelling.

Epoch AIFetched 16 Sep 2026
3.07M

Public models on Hugging Face

3,069,656 public model repositories — the only figure here that moves measurably between one refresh and the next.

Hugging FaceFetched 16 Sep 2026

On these numbers

Every figure here comes from a machine-readable source — Epoch AI publish the CSV behind their charts, and Hugging Face expose a model index. Nothing on this board is read out of an article.

The compute growth rate is mine, not Epoch’s. It is a least-squares fit over the frontier models in their dataset, and I run it rather than quote them so that the method is visible and the number moves when the data does. It lands at 4.6× a year against the 5× Epoch publish — inside the 4×–6× confidence interval they state, which is the check that the pipeline is honest.

When a source is unreachable the previous value stays, dated and flagged. A stale number that is true beats a fresh one that is invented.