AI Market Atlas · Evidence index
OpenAI 證據索引
OpenAI：保留市場快報中的可追溯指標與公開來源。
相關主題
歷史快報
引用指標
compute.openai_inference_throughput_per_watt_lower_bound基準區間 1.5–1.9 倍openai-work-septemberenergy.openai_inference_throughput_per_watt_lower_bound基準區間 1.5–1.9 倍openai-work-septembermodels.openai_business_count250 萬家企業openai-work-septembermodels.openai_frontier_output_intensity3.1 個代理工作日openai-research-septemberpulse.openai_serving_cost_reduction降低 20%openai-work-septembercompute.openai_ports_capacity約 8 IT-GWopenai-ports-pikeenergy.openai_ports_capacity約 8 IT-GWopenai-ports-pikemodels.managed_tokens1.8Topenai-pricingmodels.openai_coding_token_efficiency少 54%openai-full-stackmodels.openai_monitoring_overhead約 20%openai-cyber-pacingpulse.openai_coding_token_efficiency少 54%openai-full-stackpulse.openai_ports_capacity約 8 IT-GWopenai-ports-pikecompute.openai_ports_capacity約 8 IT-GWopenai-ports-pikeenergy.openai_ports_capacity約 8 IT-GWopenai-ports-pikemodels.managed_tokens1.8Topenai-pricingmodels.openai_monitoring_overhead約 20%openai-cyber-pacingpulse.openai_ports_capacity約 8 IT-GWopenai-ports-pikecompute.openai_ports_capacity約 8 IT-GWopenai-ports-pikeenergy.openai_ports_capacity約 8 IT-GWopenai-ports-pikemodels.managed_tokens1.8Topenai-pricingmodels.openai_monitoring_overhead約 20%openai-cyber-pacingpulse.openai_ports_capacity約 8 IT-GWopenai-ports-pikemodels.managed_tokens1.8Topenai-pricingmodels.managed_tokens1.8Topenai-pricingmodels.managed_tokens1.8Topenai-pricingmodels.managed_tokens1.8Topenai-pricingmodels.managed_tokens1.8Topenai-pricing