OpenAI has outlined its full-stack approach to artificial intelligence infrastructure, aiming to make advanced systems more capable, affordable, and accessible. According to the company, the true value of AI infrastructure lies in enabling more capable intelligence at a lower cost for a broader audience. OpenAI recently reduced prices for its GPT-5.6 Luna model by 80 percent and GPT-5.6 Terra by 20 percent. Luna now costs $0.20 per million input tokens and $1.20 per million output tokens, while Terra costs $2 and $12 respectively. Additionally, the company introduced Fast mode for GPT-5.6 Sol, offering up to 2.5 times standard processing speed at twice the price. Efficiency gains are also being driven by system-level optimizations. OpenAI reported that using GPT-5.6 Sol helped optimize production software, reducing end-to-end serving costs by 20 percent and increasing token-generation efficiency through improved speculative decoding. Furthermore, enhancements to retained reasoning and context management raised GPT-5.6 Sol's score on the public ARC-AGI-3 task set from 13.3 percent to 38.3 percent while using six times fewer output tokens, demonstrating that surrounding system design significantly impacts performance without changing the model itself. OpenAI's models now serve over one billion active users and more than two million businesses, with agentic work through Codex accounting for 99.8 percent of weekly output tokens. The company stresses that its investment decisions rely on evidence of user growth, enterprise commitments, API consumption, and technical progress, aiming to deploy the right capacity against credible demand rather than merely building the largest infrastructure.
Building abundant intelligence
OpenAI outlines its full-stack strategy for achieving AI abundance by making advanced artificial intelligence more capable, affordable, and widely accessible. The company highlights recent price reductions for GPT-5.6 models, efficiency gains from internal optimizations and system improvements, and growing global adoption across individual and enterprise users. OpenAI emphasizes a continuous feedback loop between infrastructure, research, and product usage to drive future growth and reduce serving costs.
Why it matters
The strategy illustrates how combining hardware efficiency, system-level optimizations, and aggressive pricing can lower barriers to enterprise AI adoption. It demonstrates a scalable economic model where lower costs drive broader usage, funding the next generation of AI infrastructure.