OpenAI has lowered the price of GPT-5.6 for budget-conscious customers
OpenAI has reduced the cost of using the Terra model by 20% and the Luna model by 80% / Photo: doomu / Shutterstock.com
OpenAI announced a price reduction for two of its top models designed for everyday tasks: GPT-5.6 Terra and GPT-5.6 Luna. The decision was made approximately three weeks after their public launch.
The new model family includes three models: Sol—the most powerful, Terra—the mid-range model, and Luna—the fastest. OpenAI announced that it is reducing the price of Terra by 20%—to $2 per 1 million incoming tokens and $12 per 1 million outgoing tokens. The price of Luna has dropped by 80%—to $0.2 per 1 million incoming tokens and $1.2 per 1 million outgoing tokens. The price of Sol will remain unchanged.
"Our strategy remains focused on simultaneously improving the capabilities and efficiency of our models so that each new generation of AI can do more work at a lower cost," according to a statement from OpenAI.
Context
The company is facing pressure from customers who have become more cost-conscious: corporate clients are hesitant to adopt expensive models without a clear understanding of the return on such investments, according to CNBC. For example, Amazon recently discovered that AI models can waste millions of dollars.
At the same time, OpenAI is trying to fend off competition from Chinese startups, as well as tech giants Google and Microsoft, which are promoting more cost-effective models. The AI race has been spurred by the release of the Chinese open-source AI model Kimi K3 from Moonshot AI, which, according to some industry tests, has outperformed leading American developments. Following Moonshot’s announcement, OpenAI’s main competitor—Anthropic— released a new model, Claude Opus 5, designed to handle everyday tasks.
This article was AI-translated and verified by a human editor
