Penny for your bots? AI tool calculates energy cost of chatbot prompts

Measuring AI’s energy consumption has remained largely opaque despite the technology’s growing popularity, with companies rarely disclosing the electricity demands of individual queries or models. Hugging Face engineer Julien Delavande’s new Chat UI Energy tool addresses this knowledge gap by providing real-time energy use estimates for AI conversations, making environmental impacts transparent to users and potentially establishing a new standard for energy reporting in artificial intelligence—similar to nutrition labels on food products.

The big picture: AI systems require significant energy to function despite cloud-centric marketing language that obscures their physical infrastructure requirements.

• Behind every AI query are power-hungry computers, multiple GPUs, and expansive data centers that collectively consume electricity when processing user requests.
• These energy costs partly explain why free chatbot services implement usage limits—computing is expensive for the hosting companies.

How it works: Hugging Face’s new chat interface shows real-time energy consumption estimates for user conversations with AI models.

• The tool compares energy usage across different models, tasks, and request types, revealing that reasoning-intensive prompts typically consume more energy than simple fact-finding queries.
• Users can view their consumption in technical units like Watt-hours and Joules, but also in more relatable terms such as percentage of a phone charge or equivalent driving time based on EPA data.

Real-world testing: A simple weather query about New York City demonstrated the tool’s practical applications and limitations.

• The test query consumed approximately 9.5% of a phone charge, equivalent to 45 minutes of LED bulb use, 1.21 seconds of microwave operation, or 0.15 seconds of toaster energy.
• Despite the query’s simplicity, the 90-second response time and higher-than-expected energy usage may reflect limitations from the model’s lack of internet access.

Behind the numbers: Global electricity demand projections show AI’s growing energy footprint.

• A 2024 International Energy Agency report forecasts global electricity demand increasing by 3.4% by 2026—faster than usual rates—partly driven by data center expansion.
• Berkeley Lab research predicts data centers will grow at an accelerated rate of 13% to 27% between 2023 and 2028.

What they’re saying: The Chat UI Energy team emphasizes transparency as their primary motivation for developing the tool.

• “With projects like the AI Energy Score and broader research on AI’s energy footprint, we’re pushing for transparency in the open-source community,” the creators stated.
• They envision energy usage information becoming “as visible as nutrition labels on food” in the future.

How to try it: Users can experiment with the chatbot using various open-source models.

• Available models include Google Gemma 3, Meta’s Llama 3.3, and Mistral Nemo Instruct.