What happened
OpenAI reduced GPT-5 Turbo API pricing by 40%, bringing input tokens to $1.50/M and output tokens to $6/M. The cut follows Google's Gemini 2.5 Pro pricing announcement last month and Anthropic's volume discount program. OpenAI cited inference optimization improvements and competitive pressure as drivers.
Why it matters
This is the continuation of a predictable trend: frontier model pricing drops every few months as competition heats up and inference infrastructure improves. The 40% cut is meaningful for high-volume API consumers but doesn't change the strategic calculus for most teams. If you were already building on GPT-5, your costs just went down. If you weren't, this alone isn't a reason to switch. The real story is that frontier model inference is commoditizing faster than anyone expected — the differentiation is shifting to tooling, ecosystem, and specialized capabilities.
Who should pay attention
Engineering managers budgeting for AI API costs, teams running high-volume inference workloads on OpenAI, and CFOs who need updated cost projections for AI initiatives.