Here's how much energy a single Gemini query consumes

Google has shared new insights regarding the environmental impact of its Gemini AI models. According to the company, a single text query from Gemini utilizes only 0.24 watt-hours of energy, produces 0.03 grams of carbon dioxide equivalent, and consumes 0.26 millilitres of water. To put this into perspective, the energy used is comparable to watching television for less than nine seconds. This information, released through recent technical papers and blog posts authored by Google researchers, including Chief Scientist Jeff Dean, is part of a larger initiative aimed at accurately measuring the environmental costs associated with large-scale AI operations. Previously, many assessments of AI's ecological impact were based on incomplete data, often neglecting significant factors such as idle server energy, cooling requirements, or water use. Google asserts that its refined methodology provides a more holistic view of AI inference—the process through which responses are generated by a trained model—and aims to promote standardization across the industry. In addition to these findings, Google highlights significant advancements in efficiency. Over the past year, the median energy consumption for a Gemini text prompt has decreased by a remarkable 33 times, while its carbon emissions have dropped by 44 times, all without sacrificing the quality of output. The improvements are attributed to advancements across Google's technology infrastructure, which include more efficient model designs like Mixture-of-Experts, optimized inference methods such as speculative decoding, and the development of custom Tensor Processing Units (TPUs). Notably, the latest TPU, named Ironwood, boasts an efficiency increase of 30 times compared to its inaugural TPU. Despite these efficiencies, energy demand within Google’s data centers continues to rise. In 2024, electricity consumption surged by 27 percent year-on-year, although emissions decreased by 12 percent due to a shift toward cleaner energy sources and enhanced efficiency measures. Google has committed to operating on entirely carbon-free energy and to replenishing more freshwater than it consumes in the long run. As the adoption of AI technology accelerates, Google emphasizes the need for transparency to foster accountability. "By sharing our methodology, we aim to establish a clearer benchmark for AI's true energy footprint," the company stated, reinforcing the importance of efficiency in the responsible development of AI technologies.