How does ChatGPT's per-prompt energy compare with Gemini's?

ChatGPT is listed at 0.34 Wh and Gemini at 0.24 Wh, both per company disclosures. The author prices them at 1kWh=31 yen.

What happens to AI energy use by 2030?

The IEA projects global data-center power consumption will rise from 415 TWh in 2024 to 945 TWh in 2030, roughly equal to Japan's annual consumption.

Is electricity the main constraint on AI growth?

The author says no: the bottleneck is not power but four other constraints, one of which is water. By 2030, new data-center water needs are said to equal New York City's daily use.

Large Language ModelsZenn AI/ML3m agoOct 3, 2026, 10:00 JST

AI prompt: 0.01円 — 17 billion yen a year, water is bottleneck

Send on LINE

3 Key Points

What happened
The author calculates one AI prompt costs about 0.01 yen in electricity — ChatGPT at 0.34 Wh and Gemini at 0.24 Wh, per company disclosures, at 1kWh=31 yen — and estimates all AI together at roughly 17 billion yen a year.
Why it matters
At roughly 0.01 yen per prompt, the per-use electricity cost looks negligible for an individual, so the cost picture shifts to scale and to constraints other than price.
What to watch
The 17 billion yen figure is the author's own estimate, since no organization publishes an official combined total; watch whether actual per-prompt consumption holds up as agent-style use becomes more common.

WHO IT HITSThis briefs readers tracking AI's real cost structure — energy analysts, data-center operators planning power and cooling, and finance teams modeling AI operating expenses — by putting a concrete per-prompt figure on a cost that is usually left unquantified.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

FAQ

How does ChatGPT's per-prompt energy compare with Gemini's?: ChatGPT is listed at 0.34 Wh and Gemini at 0.24 Wh, both per company disclosures. The author prices them at 1kWh=31 yen.
What happens to AI energy use by 2030?: The IEA projects global data-center power consumption will rise from 415 TWh in 2024 to 945 TWh in 2030, roughly equal to Japan's annual consumption.
Is electricity the main constraint on AI growth?: The author says no: the bottleneck is not power but four other constraints, one of which is water. By 2030, new data-center water needs are said to equal New York City's daily use.

Zenn AI/MLRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.