OpenAI cuts GPT-5.6 costs after efficiency gains
GPT-5.6 pricing drops as OpenAI cuts API costs with smarter efficiency gains See how model optimization lowers serving costs and improves value now
OpenAI said it has reduced the pricing of its GPT5.6 model family, including a large cut for the Luna variant, after internal optimization work improved efficiency and lowered serving costs.
According to the company, one of its models helped rewrite GPU code used in the system, making the models about 15% more efficient and reducing serving costs by 20%. OpenAI said those gains are being passed on to customers through lower API pricing.
The announcement comes as AI companies continue competing on price and performance. OpenAI said it wants to offer the best tradeoff between cost and intelligence across its product lineup, while rivals such as Google and openmodel developers are also pushing efficiencyfocused releases.