**[OpenAI: Over 1 Billion Active Users, Reducing AI Costs Through Full-Stack Development]** BlockBeats News, July 31 — OpenAI published an article stating that the value of AI infrastructure lies not in its scale alone but in providing more powerful intelligence to more users at lower costs. The company is building a comprehensive system encompassing infrastructure, models, platforms, and products, driving adoption growth through stronger models, and supporting next-generation research and infrastructure investments with revenue, real-world feedback, and demand data. Yesterday, OpenAI reduced the price of GPT-5.6 Luna by 80%, lowering input and output costs to $0.20 and $1.20 per million tokens, respectively. GPT-5.6 Terra saw a 20% price reduction, now priced at $2 and $12, respectively. The Fast mode of GPT-5.6 Sol can increase processing speed by up to 2.5 times without compromising intelligence levels, priced at twice the standard mode. The company stated that GPT-5.6 Sol has helped optimize model production service software, reducing end-to-end service costs by 20% and improving speculative decoding efficiency by over 15%. Currently, OpenAI models serve over 1 billion active users and more than 2 million enterprises. The goal is not simply to build more computational power or larger models but to deploy appropriate capacity based on reliable demand, making useful intelligence stronger, cheaper, and accessible to a broader range of users. [Original Link]