OpenAI announced Thursday it is reducing the price of two models in its GPT-5.6 family — cutting the cost of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% — roughly three weeks after the models launched, according to CNBC.
The cuts come as businesses grow wary of rising AI bills and cheaper Chinese models intensify competition
SOPA Images / Getty Images
OpenAI announced Thursday it is reducing the price of two models in its GPT-5.6 family — cutting the cost of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% — roughly three weeks after the models launched, according to CNBC.
Luna, the fastest and lowest-cost version in the GPT-5.6 lineup, is now priced at 20 cents per million input tokens and $1.20 per million output tokens, a reduction from its previous rates of $1 and $6. Terra, the mid-tier model, will see its input token rate fall to $2 per million and output token rate fall to $12 per million, compared with the prior $2.50 and $15. The pricing for Sol, the most powerful model in the family, remains unchanged, the company said.
OpenAI attributed the reductions to efficiency gains made during internal development of GPT-5.6, including the model's ability to rewrite and optimize production code and improve token generation, the company said. Those improvements reduced the end-to-end cost of serving the model by 20% and increased token-generation efficiency by more than 15%.
The cuts arrive as enterprise customers have grown reluctant to authorize large AI expenditures without a clear picture of returns, according to Reuters. Cheaper Chinese open-weight models have sharpened the competitive environment, pushing OpenAI and rival Anthropic to make the case that their products are worth the premium, according to Axios.
The broader backdrop has been building for some time. Enterprise customers have been cutting AI spending and demanding clearer returns on investment, with some companies switching to cheaper alternatives. Ride-sharing company Uber $UBER burned through its entire annual AI budget in four months before introducing spending tiers on some tools. Both OpenAI and Anthropic have also filed confidential IPO prospectuses with the Securities and Exchange Commission, adding pressure to demonstrate durable business models ahead of potential public listings.
OpenAI had been considering price reductions as recently as last month. CEO Sam Altman called costs "a huge issue" at a recent event, saying the company would find ways to help customers get more value for less spending.
By comparison, Anthropic's mid-tier Claude Sonnet 4.6 carries a price tag of $3 per million input tokens and $15 per million output tokens, putting it above what Terra now costs, according to Reuters. Microsoft $MSFT and Google $GOOGL have each been expanding lower-cost model offerings aimed at business customers, according to CNBC.
Along with the price reductions, OpenAI said it is introducing a Fast mode in its API for GPT-5.6 Sol, delivering up to 2.5 times faster speeds than standard processing at twice the price, replacing its previous Priority Processing offering.
Join 500,000+ readers who start their day with Quartz.
By subscribing, you agree to our Terms of Service and Privacy Policy.