Generative AI is billed per token, yet tokens are not equally expensive. Each token enters the KV-cache and raises the cost of those that follow, so energy grows non-linearly along a sequence whose to
A four-parameter analytical law predicts LLM inference energy across architectures, revealing quadratic prefill costs, GQA savings, and super-linear growth from reasoning budgets and agentic serving.
Generative AI is billed per token, yet tokens are not equally expensive. Each token enters the KV-cache and raises the cost of those that follow, so energy grows non-linearly along a sequence whose tokens look identical. No analytical model states what a token costs, and the most accurate SOTA model is a fit, not a law: it absorbs architecture into fitted coefficients, requiring per-LLM fitting. We derive closed-form expressions of the computational and memory-access complexity of LLM inference in terms of architectural hyperparameters, covering MHA and GQA, and compare a family of analytical models of increasing granularity. From these, we obtain the SIMPLE Inference Model to Predict LLMs…
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
The SIMPLE Law for Energy Cost of LLM Inference was first seen on Hugging Face · Daily papers on Nov 5, 2026. Since then it has appeared in 37 checks on 4 different days. Its best position was #19 of 30, reached on Nov 6, 21:28 UTC. It is #23 right now, slipping from #22 to #23 over the last 6 checks.
Position on Hugging Face · Daily papers at every check (lower is better).
Best rank per day (UTC).
▲ as reported by Hugging Face.
Archived daily, weekly and monthly lists that include this entry.
Yes. As of Nov 8, 17:28 UTC it is #23, slipping from #22 to #23 over the last 6 checks.
#19 on Hugging Face · Daily papers, on Nov 6, 21:28 UTC. Trendoday checks this list every 120 minutes.
It was first seen on Nov 5, 2026 and has been on the list on 4 days (37 checks) since.