Research

Measuring the Inference Energy Efficiency of Open-Weight LLMs across Benchmarks, Reasoning Modes, and Hardware

Single source1 source14 Oct 2026

arXiv:2610.14389v1 Announce Type: new Abstract: Benchmark scores tell us what an LLM can do, but not how much energy is required to deliver that capability.

Reported by 1 source

arxiv.orghttps://arxiv.org/abs/2610.14389

← Back to all developments · Signals are a development reported by more than one source. See the methodology.