The Energy Footprint of LLM-Based Environmental Analysis: LLMs and Domain Products

TL;DR

This paper evaluates the energy consumption of domain-specific RAG workflows versus generic LLMs in climate analysis, highlighting how design choices impact energy use and output quality.

Contribution

It provides the first detailed assessment of energy use in domain-specific LLM applications, emphasizing the influence of workflow design on energy footprint and response quality.

Findings

Agentic pipelines increase energy consumption significantly.

Energy use varies with access location and time of day.

Design choices in RAG systems affect both energy footprint and output quality.

Abstract

As large language models (LLMs) are increasingly used in domain-specific applications, including climate change and environmental research, understanding their energy footprint has become an important concern. The growing adoption of retrieval-augmented (RAG) systems for climate-domain specific analysis raises a key question: how does the energy consumption of domain-specific RAG workflows compare with that of direct generic LLM usage? Prior research has focused on standalone model calls or coarse token-based estimates, while leaving the energy implications of deployed application workflows insufficiently understood. In this paper, we assess the inference-time energy consumption of two LLM-based climate analysis chatbots (ChatNetZero and ChatNDC) compared to the generic GPT-4o-mini model. We estimate energy use under actual user queries by decomposing each workflow into retrieval, generation, and hallucination-checking components. We also test across different times of day and geographic access locations. Our results show that the energy consumption of domain-specific RAG systems depends strongly on their design. More agentic pipelines substantially increase inference-time energy use, particularly when used for additional accuracy or verification checks, although they may not yield proportional gains in response quality. While more research is needed to further test these initial findings more robustly across models, environments and prompting structures, this study provides a new understanding on how the design of domain-specific LLM products affects both the energy footprint and quality of output.