Verified65% confidenceFactExact time
It is estimated that training GPT-3 once emits approximately 552 metric tons of CO2 equivalent.
3
Sources
65%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
UnverifiedIt is estimated that training GPT-3 produced carbon emissions roughly equivalent to 125 round-trip flights across the United States.78% similarVerifiedIt is estimated that training a large language model of GPT-3's scale requires approximately 1,287 megawatt-hours of electricity.70% similarVerifiedResearch shows that a single training run of GPT-3 consumes approximately 700,000 liters of fresh water, though this figure is subject to considerable controversy due to varying system boundary definitions.69% similarUnverifiedGPT-3训练耗资约450万美元67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/573245API
curl https://kongchang.com/api/v1/knowledge/claims/573245MCP
get_claim(id=573245)