# How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference

arXiv: https://arxiv.org/abs/2505.09598 · HTML: https://mlsanity.com/paper/2505.09598/

## Key facts

- Cited by 76 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 74 references that are in the map.
- First posted on arXiv 14 May 2025.

## Builds on (most cited first, 16 of 16)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [Carbon Emissions and Large Neural Network Training](https://mlsanity.com/paper/2104.10350/) (2021)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Humanity's Last Exam](https://mlsanity.com/paper/2501.14249/) (2025)
- [Power Hungry Processing: Watts Driving the Cost of AI Deployment?](https://mlsanity.com/paper/2311.16863/) (2023)
- [SciCode: A Research Coding Benchmark Curated by Scientists](https://mlsanity.com/paper/2407.13168/) (2024)
- [From Words to Watts: Benchmarking the Energy Costs of Large Language Model Inference](https://mlsanity.com/paper/2310.03003/) (2023)
- [LLM-Inference-Bench: Inference Benchmarking of Large Language Models on AI Accelerators](https://mlsanity.com/paper/2411.00136/) (2024)
- [The Price of Prompting: Profiling Energy Use in Large Language Models Inference](https://mlsanity.com/paper/2407.16893/) (2024)
- [Double-Exponential Increases in Inference Energy: The Cost of the Race for Accuracy](https://mlsanity.com/paper/2412.09731/) (2024)
- [Empirical Measurements of AI Training Power Demand on a GPU-Accelerated Node](https://mlsanity.com/paper/2412.08602/) (2024)

## Built on by (most cited first, 30 of 76)

- [TokenPowerBench: Benchmarking the Power Consumption of LLM Inference](https://mlsanity.com/paper/2512.03024/) (2025)
- [Energy Use of AI Inference, Efficiency Pathways, and Test-Time Scaling](https://mlsanity.com/paper/2509.20241/) (2025)
- [OckBench: Measuring the Efficiency of LLM Reasoning](https://mlsanity.com/paper/2511.05722/) (2025)
- [EET: Experience-Driven Early Termination for Cost-Efficient Software Engineering Agents](https://mlsanity.com/paper/2601.05777/) (2026)
- [Otters: An Energy-Efficient SpikingTransformer via Optical Time-to-First-Spike Encoding](https://mlsanity.com/paper/2509.18968/) (2025)
- [Do Chatbot LLMs Talk Too Much? The YapBench Benchmark](https://mlsanity.com/paper/2601.00624/) (2026)
- [ECVL-ROUTER: Scenario-Aware Routing for Vision-Language Models](https://mlsanity.com/paper/2510.27256/) (2025)
- [Edge Deployment of Small Language Models, a comprehensive comparison of CPU, GPU and NPU backends](https://mlsanity.com/paper/2511.22334/) (2025)
- [Prompt Injection Mitigation with Agentic AI, Nested Learning, and AI Sustainability via Semantic Caching](https://mlsanity.com/paper/2601.13186/) (2026)
- [AI Benchmark Democratization and Carpentry](https://mlsanity.com/paper/2512.11588/) (2025)
- [This Is Taking Too Long -- Investigating Time as a Proxy for Energy Consumption of LLMs](https://mlsanity.com/paper/2603.15699/) (2026)
- [Stable FP4 Training via Transposition-Invariant Block Quantization](https://mlsanity.com/paper/2607.24953/) (2026)
- [Accurate and Energy Efficient: Local Retrieval-Augmented Generation Models Outperform Commercial Large Language Models in Medical Tasks](https://mlsanity.com/paper/2506.20009/) (2025)
- [Walking the Tightrope of LLMs for Software Development: A Practitioners' Perspective](https://mlsanity.com/paper/2511.06428/) (2025)
- [The Energy Footprint of LLM-Based Environmental Analysis: LLMs and Domain Products](https://mlsanity.com/paper/2604.00053/) (2026)
- [Llamas on the Web: Memory-Efficient, Performance-Portable, and Multi-Precision LLM Inference with WebGPU](https://mlsanity.com/paper/2605.20706/) (2026)
- [Energy per Successful Goal: Goal-Level Energy Accounting for Agentic AI Systems](https://mlsanity.com/paper/2605.22883/) (2026)
- [Phishing Detection in the Gen-AI Era: Quantized LLMs vs Classical Models](https://mlsanity.com/paper/2507.07406/) (2025)
- [Learning the Topic, Not the Language: How LLMs Classify Online Immigration Discourse Across Languages](https://mlsanity.com/paper/2508.06435/) (2025)
- [Dissecting Transformers: A CLEAR Perspective towards Green AI](https://mlsanity.com/paper/2510.02810/) (2025)
- [Performance Trade-offs of Optimizing Small Language Models for E-Commerce](https://mlsanity.com/paper/2510.21970/) (2025)
- [Analysing Environmental Efficiency in AI for X-Ray Diagnosis](https://mlsanity.com/paper/2511.07436/) (2025)
- [David vs. Goliath: Can Small Models Win Big with Agentic AI in Hardware Design?](https://mlsanity.com/paper/2512.05073/) (2025)
- [Addressing Climate Action Misperceptions with Generative AI](https://mlsanity.com/paper/2602.22564/) (2026)
- [Scaling Quantum Machine Learning without Tricks: Full-Resolution and Diverse Image Generation](https://mlsanity.com/paper/2603.00233/) (2026)
- [Multilinguality at the Edge: Developing Language Models for the Global South](https://mlsanity.com/paper/2604.21637/) (2026)
- [Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities](https://mlsanity.com/paper/2605.04127/) (2026)
- [MARLIN: Multi-Agent Game-Theoretic Reinforcement Learning for Sustainable LLM Inference in Cloud Datacenters](https://mlsanity.com/paper/2605.13496/) (2026)
- [M"OVE: A Holistic LLM Benchmark for the German Public Sector](https://mlsanity.com/paper/2606.13111/) (2026)
- [Attention to Detail: Evaluating Energy, Performance, and Accuracy Trade-offs Across vLLM Configurations](https://mlsanity.com/paper/2607.09172/) (2026)

## Abstract (arXiv)

This paper introduces an infrastructure-aware benchmarking framework for quantifying the environmental footprint of LLM inference across 30 state-of-the-art models in commercial datacenters. The framework combines public API performance data with company-specific environmental multipliers and statistical inference of hardware configurations. We additionally utilize cross-efficiency Data Envelopment Analysis (DEA) to rank models by performance relative to environmental cost and provide a dynamically updated dashboard that visualizes model-level energy, water, and carbon metrics. Results show the most energy-intensive models exceed 29 Wh per long prompt, over 65 times the most efficient systems. Even a 0.42 Wh short query, when scaled to 700M queries/day, aggregates to annual electricity comparable to 35{,}000 U.S. homes, evaporative freshwater equal to the annual drinking needs of 1.2M people, and carbon emissions requiring a Chicago-sized forest to offset. These findings highlight a growing paradox: as AI becomes cheaper and faster, global adoption drives disproportionate resource consumption. Our methodology offers a standardized, empirically grounded basis for sustainability benchmarking and accountability in AI deployment.