ทันเอไอ

OpenAI cuts API prices for GPT-5.6 Terra by 20% and Luna by up to 80%

OpenAI has announced price cuts for API usage of its mid-tier Terra model by 20% and its small Luna model by up to 80%, citing system efficiency improvements that lowered operating costs

OpenAI has announced price reductions for API usage of two new models in the GPT-5.6 series, effective July 30. Terra, the mid-tier model, sees a 20% price cut from its launch pricing, while Luna, the speed-focused small model, drops by as much as 80%.

Terra's new pricing is $2 per 1 million input tokens and $12 per 1 million output tokens, down from $2.50 and $15 respectively. Luna now costs just $0.20 per 1 million input tokens and $1.20 per 1 million output tokens, down from $1 and $6 — an aggressive price cut for a budget-tier model.

Sol, the most powerful model in the series, keeps its original pricing at $5/$30 per 1 million input/output tokens, but gains a new addition: a Fast mode claimed to process 2.5 times faster, priced at twice the standard rate, aimed at workloads requiring rapid response times.

OpenAI explained that the price cuts stem from internal efficiency improvements. Its teams optimized GPU software and model-serving code automatically, reducing serving costs by 20% and improving token generation efficiency by more than 15%. The company chose to pass these savings on to customers in the form of lower prices rather than pocketing them as extra profit.

This pricing move reflects a broader industry trend in which AI model usage costs keep falling due to competition and advancing technology, enabling developers and businesses to deploy AI in production at lower budgets. Luna in particular, with its 80% price drop, could become the go-to option for workloads requiring massive-scale processing.