How much energy does ChatGPT use?

Numbers match the citations in the article. A working link isn't proof that a page supports a claim; check the quoted passage and date.

1. 1

   A Locally Executable AI System for Improving Preoperative Patient Communication: Multidomain Clinical Evaluation.

   JMIR medical informatics (Sato et al.)Published Jul 21, 2026Checked Oct 10, 2026

   “This performance was not statistically different from that of ChatGPT (GPT-4o), which had 6 out of 400 (1.5%) errors (McNemar test with Holm adjustment; P>.99). Sustainability measurements showed approximately 2.23 mWh per request (latency≈0.10 s; video RAM≈2.2 GiB average, ≈2.5 GiB peak) for the nongenerative clinical path vs approximately 168.27 mWh (latency≈8.51 s; video RAM≈13.3 GiB average, ≈14.0 GiB peak) for small language model small talk-approximately a 75-fold higher energy footprint per reply for the generative path.ConclusionsHigh-precision, nongenerative clinical support is feasible using local, low-cost hardware without cloud dependence. By decoupling clinical information retrieval from generative chitchat, LENOHA enhances safety, preserves privacy, and markedly reduces energy use, offering a practical blueprint for sustainable and equitable medical AI deployment across diverse care settings.”

2. 2

   Green prompt engineering for sustainable generative AI.

   Environmental science and ecotechnology (Podder et al.)Published Mar 13, 2026Checked Oct 10, 2026

   “However, substantial computational costs and energy footprint of prompt inferencing process remain critical challenges while building generative AI applications. The energy efficiency of LLM inferences is particularly impacted by suboptimal prompts, which may require multiple iterations, thereby escalating energy consumption and the associated carbon footprint. To address these challenges, we propose a series of practices and guidelines designed to enhance the likelihood of obtaining desired responses from LLMs with minimal reiterations. Empirical evaluation demonstrates that, across a range of LLMs and test scenarios, energy consumption and corresponding operational greenhouse gas emissions were reduced by 32-48% when best practices were applied. Drawing upon these insights, our proposed best practices can be seamlessly integrated into the design frameworks of generative AI applications, thereby enhancing the energy efficiency of prompt inferencing. By addressing the challenge of establishing a cohesive framework for energy-efficient prompt design and inferencing, this paper advocates for the sustainable and effective deployment of generative AI technologies.”

3. 3

   Carbon Reporting Practices in the NHS: Emissions and Omissions Relating to Artificial Intelligence.

   Journal of medical Internet research (Reynolds)Published Oct 27, 2025Checked Oct 10, 2026

   “First, a lack of granularity provides averages that can obscure the extreme energy intensity of certain AI workloads. Second, life-cycle emissions from specialized hardware (eg, graphics processing units) are often excluded unless trusts own the equipment, ignoring upstream manufacturing impacts. Third, widespread use of unprocured generative AI tools is unmeasured; extrapolating general practice survey data suggests that ChatGPT queries alone could release ≈ 349t CO₂e per year in primary care. To close these gaps, we propose three potential ways to help reduce these reporting gaps: (1) AI-specific carbon disclosure clauses in vendor contracts, (2) inclusion of cradle-to-grave emission factors for AI hardware in Scope 3 reporting, and (3) lightweight monitoring of external AI traffic (while recognizing potential ethical issues with this). Implementing these measures would give health care leaders a more accurate baseline against which to judge whether AI supports or undermines the NHS net-zero target.”

4. 4

   Quantifying and Disclosing the Environmental Footprint of AI in Research: Life Cycle-Informed Framework and Open-Access Calculator Development Study.

   JMIR AI (Mozafari et al.)Published Aug 18, 2026Checked Oct 10, 2026

   “In the small laboratory scenario, optimization reduced total emissions from approximately 0.07 kg carbon dioxide equivalents (CO2e) to 0.05 kg CO2e and improved the proposed label from C to B. In the midsized cluster scenario, carbon-aware scheduling reduced emissions from approximately 665 kg CO2e/month to 450 kg CO2e/month-a 32% reduction. In the large-scale scenario, shifting to renewable-backed, lower-PUE infrastructure reduced operational emissions by approximately 79% while increasing the importance of water-carbon trade-off reporting.ConclusionsThe calculator provides a practical and transparent method for reporting AI environmental footprints using auditable parameters and publication-ready outputs. Routine disclosure of energy use, emissions, water use, hardware assumptions, and regional context can improve reproducibility and support more equitable and sustainable AI research evaluation.”

5. 5

   Large language model (Wikipedia)

   WikipediaPublished Oct 10, 2026Checked Oct 10, 2026

   “A large language model (LLM) is an AI model (typically a neural network) trained on a vast amount of text for natural language processing tasks, especially language generation. LLMs can typically generate, summarize, translate, and analyze text in many contexts. They are the basis for many modern chatbots, such as ChatGPT, Claude, Gemini, Grok, Qwen and DeepSeek. LLMs are typically based on the transformer neural network architecture. Generative pre-trained transformers (GPTs) are a type of LLM that is pre-trained to predict the next word. GPTs are then often fine-tuned to follow instructions and to behave as assistants. Biased or inaccurate training data can make an LLM's output less reliable. A software harness can provide LLMs with memory and access to tools such as web search. Benchmark evaluations for LLMs attempt to measure model reasoning, factual accuracy, alignment, and safety.”