ChatgptOpenai's chatbotpot platform may not be hungry as it once believed. According to the new study, Champope is based on the majority of AI models that answer the chatgpt and queries.
One Currently analysis The non-profit AI Research Institute has tried to calculate how much energy it is. One Mostly reference stat Chatgpt is about 3 hours to answer a question or Google Search for about 10 times the Watt-Child Power.
Gun is believed that it is extremely sensitive.
Using the latest default model of OpenAIAIA; GPT-4OAs a reference, George is less than 0.3 Watt-Hours 0.3 hours – Home Appliances.
“Energy consumption uses normal users using regular users using regular equipment or not a big deal compared to the driver.
AI's energy consumption – the widespread talking and environmental impact – AI companies are controversial debates to see the steady expansion of their infrastructure. A group of more than 100 organizations last week Published open letter New AI Data Centers call AI industry and regulations to ensure AI data centers to rely on natural resources and non-force resources.
Techcrunch said that he was motivated by the characteristics of his appearance, as the former researchers were researching his analysis. For example, the writer of the report that entered the Watt-hung forecasts used the openaui to open its models. Glarai is supposed to be used.

“In the coming years, many people have found a lot of public talks that the AI will be eating a lot of energy,” he said.
The 0.3 Watt-Houaung picture of Epchoch is approximately. Openai does not publish the details required to make a definite calculation.
The analysis does not consider additional energy expenses such as imagery or input processing, such as chatgpt features. For example, queries with the long input “Charget Querties – for long files – querrties – querrties – more electricity supply more than the usual value.
But he hopes that it will increase baseline chatgpt energy consumption.
“The Ai will be more advanced. This future AI will need a lot of energy. This future AI can use it more and more.
While there Notable achievements Ai of Ai is a major factor of AI in recent months in recent months. The AI data centers in the next two years may require almost all 68 GW power capacity (68 GW). According to a compared report. In the year 2030, the report was estimated to be able to demand the production of energy production with electricity production in the eight-GW nuclear reactor.
ChatgT alone is very large and widely and its server is similarly demanded. Openai with investment partners, with plans Billions of dollars in the new AI Data Center Project Used Within the next few years.
Openai's attention – with the rest of the other AI independing -, they are generally changing the most effective models that are generally more efficiently capable for the work they can accomplish. Contradict models against models such as GPT-4O. The investigations immediately responded.
“Reasoning models will increase the work that older people cannot do older people.
Openai releases more energy efficient rational patterns o3-mini. However, at least this turning point in this turns, this is not possible that it is impossible for energy needs to meet energy needs due to reasoning and growing ai use of high performance.
They are worried about the Ai Energy FootprintPrints.
“Openai's GPT-4o-mini can be used,” he said.