OpenAI Cuts GPT-5.6 Luna Pricing by 80%, Launches Fast Mode Across Model Family

OpenAI announced a major pricing overhaul for its GPT-5.6 model family on July 30, 2026, slashing Luna's cost by 80% to $0.20 per million input tokens and $1.20 per million output tokens. The mid-tier Terra model also received a 20% price reduction, while the flagship Sol model retained its existing price. OpenAI simultaneously rebranded its Priority Processing tier as Fast Mode, now available across all three models at double the standard rate but delivering up to 2.5x faster response speeds. The company attributed the cuts to backend infrastructure improvements, including a roughly 20% reduction in serving costs and over 15% gains in token generation efficiency. OpenAI positioned the changes as enabling businesses to deploy cost-effective agentic workflows at scale, with Luna targeting high-volume tasks and Sol reserved for complex reasoning work.