OpenAI cuts prices for GPT-5.6 Luna and Terra models, says optimization of its own infrastructure by Sol made this possible
From 30 July, OpenAI reduced the price of GPT-5.6 Luna by 80 % (to 0.20 $/1.20 $ per million tokens) and Terra by 20 %. According to the company, the cuts stem from savings achieved by optimizing infrastructure with Sol; commentators point to pressure from low-cost competitors.
From 30 July 2026, OpenAI reduced usage prices for GPT-5.6 Luna by 80 % and Terra by 20 %; the price of Sol remains unchanged. Luna now costs 0.20 dollars per million input tokens and 1.20 dollars per million output tokens, while Terra costs 2 dollars and 12 dollars per million tokens. All three models are available through ChatGPT Work, Codex and OpenAI API. According to OpenAI, the performance of Luna matches the best models from a year ago, but a task that cost a dollar then now costs approximately 6 cents on Luna and runs almost nine times faster.
According to OpenAI, the price cuts were made possible by GPT-5.6 Sol helping to make the internal operational infrastructure of the company more efficient. The company says Sol independently rewrote and optimized production kernels for model computation, including in Triton and Gluon, languages maintained by OpenAI itself. According to the company, these changes reduced overall operating costs by 20 %, and speculative decoding improved token generation speed by more than 15 %.
Independent commentator Simon Willison points out that, following the price cut, Luna is cheaper than Gemini 3.1 Flash-Lite from Google (around 0.25 dollars/1.50 dollars per million tokens according to his comparison) and roughly five times cheaper for input than the cheapest model from Anthropic, Claude Haiku 4.5 (1 dollar/5 dollars per million tokens), although their prices were previously comparable. Willison says this prompted him to move his own demo application from Gemini 3.1 Flash-Lite to Luna.
The Decoder links the move to growing price pressure in the AI market, including low-cost Chinese providers and the fact that Microsoft is now actively offering its MAI models as a cheaper alternative to models from OpenAI. The publication warns that, in its view, such a price war could weaken revenue at leading AI labs whose finances depend on extensive infrastructure investment.
Why it matters
The lower price per million tokens for the low-cost Luna model means that companies and individuals can run AI features (chatbots, agents, batch text processing) at significantly lower costs than before, changing the calculation of whether deploying AI at a larger scale is worthwhile. It also signals an intensifying price war between OpenAI, Google, Anthropic and Microsoft, which could give developers and companies cheaper access to models while increasing pressure on provider margins.
Two audiences, two different impacts
What this means
For individuals
Anyone using OpenAI API for their own projects or tools with a low-cost model (Luna) will pay significantly less for the same work than before, as illustrated by Simon Willison, who switched from Gemini to Luna because of the price.
For a business
Companies running products on cheaper GPT-5.6 models can reduce inference costs by as much as tens of percent or consider switching from competing providers (Google, Anthropic), intensifying price pressure across the AI industry.
StrategyCheck the original
Event sources
confirmed by 2 independent sources · 3 publishers, 2 independent. We count feeds from the same owner only once.