Short answer: a current median text prompt is reported at 0.24 watt-hours (Wh) for Gemini, while an average ChatGPT query is reported at 0.34 Wh. A historical estimate put a conventional Google search at about 0.3 Wh. Those figures are close enough that the familiar claim that every AI query uses 10 times as much electricity as a search is not a reliable rule today.
The comparison depends on the model, prompt and response length, whether reasoning or tool calls are involved, what the measurement includes, and when it was made. Older systems were estimated at 2.9 Wh per ChatGPT request, which is why early headlines produced a much larger ratio.
As an Amazon Associate I earn from qualifying purchases.
At-a-glance comparison
Watt-hours measure electricity, not computing time or carbon emissions. One watt-hour is the energy used by a one-watt device running for one hour. The figures below are estimates for different workloads and accounting boundaries, so they should be read as order-of-magnitude comparisons rather than readings from one common meter.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minute| Service or estimate | Date | Workload | Energy boundary and method | Reported electricity | How to interpret it |
|---|---|---|---|---|---|
| Gemini | 2025 | Median text prompt | Google measurement including operational data-center energy; company methodology | 0.24 Wh | Current median, so half of the measured prompts were lower and half higher in that distribution |
| ChatGPT | 2025 | Average query | Company-reported figure from OpenAI, recorded by the International Energy Agency (IEA) | 0.34 Wh | An average reported by the provider, not an independently audited universal value |
| Frontier-scale model | 2025 | Median text interaction | Microsoft Research bottom-up estimate summarized by the IEA | 0.34 Wh | Modeled estimate rather than a direct meter reading |
| ChatGPT | 2023 | Request in an early estimate | Estimate reviewed in Joule; assumptions reflect earlier, less efficient systems | 2.9 Wh | Historical value behind many “10 times a search” stories |
| Conventional Google search | Historical estimate summarized in 2024 | Standard, non-generative search | Technical-report estimate; not a current audited benchmark | About 0.3 Wh | Useful baseline, but modern search pages may do more than this original workload |
| Associated Press modeled comparison | 2025 | Simple and complex AI prompts versus an AI-free Google search | Modeled examples, not fixed measurements for every query | About 23 times and 210 times the search baseline | Illustrates how workload complexity can dominate the result |
The current disclosed text-prompt values, 0.24–0.34 Wh, sit near the historical 0.3 Wh search estimate. That numerical closeness does not prove that a particular prompt and a particular search consume the same amount: the studies use different dates, systems and boundaries.
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
Why there is no single “AI query” number
Prompt and response length
A short question that produces a short answer requires fewer tokens and less computation than a long document analysis. Supplying a large conversation history or several files makes the request heavier even when the final answer is brief. A median value describes a distribution of ordinary prompts; it is not a cap or a promise for every request.
Model size and routing
Services may route easy requests to a smaller model and difficult requests to a larger one. The user-facing product name does not reveal which model handled a specific turn. Two prompts that look similar in the interface can therefore have different electricity use.
Reasoning, agents and tool calls
Some systems generate hidden intermediate steps, retry a response, browse, run code or call other services. Each additional model pass adds computation. A request that triggers several steps should not be compared with a one-pass text completion using the same per-query label.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Hardware utilization and cooling
Electricity depends on accelerator efficiency, how fully the hardware is utilized, and the data center’s cooling and power-delivery overhead. Google’s Gemini figure explicitly includes operational data-center energy, while other estimates may emphasize model-computation energy or use a different facility boundary. These choices can move the result even when the underlying model work is similar.
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
Location and accounting period
Data centers operate on different power systems and efficiency levels. A measurement from one fleet and year cannot automatically be transferred to another. The 2025 figures reflect newer hardware and software than the 2023 estimate, which is one reason the numbers differ so sharply.
What the “10 times more than Google” claim actually means
The headline came from dividing the early 2.9 Wh ChatGPT estimate by the roughly 0.3 Wh historical estimate for a conventional Google search. The arithmetic is about 9.7, commonly rounded to 10. It was a comparison of a 2023-era AI request with an older search baseline, not a timeless physical constant.
Using the newer disclosed values changes the picture. The 0.24 Wh Gemini median is numerically below the 0.3 Wh historical search estimate, and the 0.34 Wh ChatGPT average is slightly above it. Because the measurements are not produced under one standardized protocol, it would be misleading to turn those differences into a precise percentage advantage for either service.
The Associated Press examples show why a single ratio can still be useful as a warning. Its model estimated a simple AI prompt at about 23 times the energy of an AI-free Google search and a complex prompt at about 210 times. Those are modeled scenarios, not readings that apply to every simple or complex prompt, but they demonstrate that adding substantial generation or reasoning can overwhelm the baseline comparison.
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
What counts as a web search?
The 0.3 Wh figure refers to a historical estimate for a conventional search. A modern results page can include advertising auctions, richer ranking, personalization, media previews and an AI-generated summary. Those components change the workload, so a current search session is not necessarily equivalent to the old “one search” baseline.
There is no universal published meter reading for an AI overview embedded in a search page. An overview that invokes a generative model will require additional computation beyond retrieving and ranking links, but its electricity use depends on the model, context, caching and the amount of text generated. The available modeled comparisons support the direction of that effect without establishing one fixed number for all AI-overview pages.
Which tasks use the most electricity?
Short text questions
A brief prompt with a brief answer is the lightest case represented by the current 0.24–0.34 Wh figures. Even here, the values are medians or averages, so unusually long inputs, peak demand or a different model can produce a higher result.
Long reasoning and document work
Multi-step reasoning, long context windows, spreadsheet analysis and repeated revisions require more token generation and often more model passes. The modeled “complex prompt” result from the Associated Press is intended to capture this kind of heavier workload.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
Images, audio and video
Generating or transforming images, transcribing or synthesizing audio, and creating video involve substantially more computation than a short text completion. The text-prompt figures should not be reused for those modalities; the available data does not establish a universal Wh value for them.
Agentic workflows
An agent that searches, opens pages, extracts information, writes code, executes a tool and checks its own result may issue many model calls for one user instruction. Counting that as “one query” hides the electricity consumed by the individual steps.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Individual use versus system-wide demand
For one person, a fraction of a watt-hour is small. Arithmetic using the Gemini median illustrates the scale: 100 such prompts would total 24 Wh, and 1,000 would total 240 Wh, or 0.24 kilowatt-hours. These are simple multiplications of the reported median, not measurements of a household’s bill.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The infrastructure question is different because services handle enormous volumes. The IEA estimates that running all conventional internet searches as simple AI text queries would require less than 4 terawatt-hours per year, which it describes as less than 1% of current data-center electricity consumption. This is a scenario estimate, not a forecast that every search will be replaced by AI.
The IEA also reports that data-center electricity demand grew 17% in 2025 and expects more intensive AI workloads to influence future demand. Aggregate growth can therefore be material even when the energy attached to one ordinary text prompt remains a fraction of a watt-hour.
How to estimate the electricity for your own usage
You can make a transparent, rough estimate when you know how many requests you send and which published figure is closest to your workload.
- Count requests over the period. Use prompts or completed interactions, not just days you opened the app.
- Classify the workload. Separate short text from long reasoning, file analysis, image generation, audio, video and agentic tasks.
- Choose a matching estimate. For ordinary text, 0.24 Wh (Gemini median) or 0.34 Wh (reported ChatGPT average) are current reference points. Do not apply them to media generation.
- Multiply and convert. Requests × watt-hours per request gives watt-hours; divide by 1,000 for kilowatt-hours.
- Label the result as an estimate. State the date, workload and source figure, because changing any of those assumptions changes the answer.
For example, multiplying 500 ordinary text prompts by 0.34 Wh gives 170 Wh, or 0.17 kWh. That calculation does not include the electricity used by your phone or computer, network equipment, or the embodied energy of manufacturing the hardware. It also cannot reveal the provider’s exact data-center allocation for your account.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →How to reduce unnecessary AI electricity use
- Choose the least intensive tool that solves the task. A conventional search may be sufficient when you need one known fact or a link rather than a generated explanation.
- Ask a focused question. Clear constraints can reduce back-and-forth retries and oversized answers.
- Reuse a good result. Saving or quoting an answer for later avoids regenerating the same text repeatedly.
- Avoid needless high-cost modalities. Do not request an image, audio rendering or video when plain text will do.
- Batch related work carefully. Combining closely related questions can avoid repeated context, but attaching very large files or asking for extensive reasoning may increase one request’s cost; there is no universal winner.
- Turn off or limit autonomous steps when they are unnecessary. Browsing, code execution and multi-step agents add calls beyond the initial prompt.
These practices reduce avoidable computation, but they do not produce a precise personal carbon or electricity audit. The provider’s hardware, utilization, cooling and grid mix remain outside the user’s control.
How to read future energy claims
Before accepting a comparison, check six items:
- Date: newer hardware can make an old estimate obsolete.
- Workload: identify whether it is a short text prompt, long reasoning task, image, audio, video or a multi-step agent.
- Statistic: a median, average, range and worst case answer different questions.
- Boundary: determine whether the figure covers model computation only, operational data-center energy, or a wider system.
- Method: distinguish direct company measurement, a bottom-up model and a historical estimate.
- Search definition: verify whether “web search” means a bare link query or a modern page with ads, media and generative summaries.
If any of these details are missing, treat the number as a rough illustration rather than a universal conversion factor.
Bottom line
For current, ordinary text use, the best disclosed reference points are about 0.24 Wh for a median Gemini prompt and 0.34 Wh for an average ChatGPT query. Those values are in the same broad range as the historical 0.3 Wh estimate for a conventional search. The older 2.9 Wh ChatGPT estimate explains the popular 10× claim, while modeled complex tasks can produce much larger ratios. The honest answer is therefore conditional: a short AI question is not automatically ten times as electricity-intensive as a search, but long reasoning, media generation and multi-step agents can be far more demanding.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
