Advancing the price-performance frontier with GPT-5.6

What happened

6⁠ helped make itself more efficient to run⁠. 6 Sol⁠ in the API. Together, these updates help customers get more from every dollar they invest in AI and move faster when time matters. 6 Terra, our balanced model for everyday work, will cost 20% less. These lower prices for Luna and Terra are also reflected in how usage is counted against paid subscriptions when using Codex and ChatGPT Work.

Luna gives businesses a far more cost-effective way to handle high-volume work at very high levels of quality. It can use tools and complete multi-step workflows, making a broader range of AI applications practical to run at scale. Making advanced intelligence more abundant and affordable is central to OpenAI’s mission to ensure AGI benefits all of humanity. These changes put that commitment into practice.

They reflect years of improvements in how our models are built, served, and put to work. We’re also introducing Fast mode in the API, which replaces our Priority Processing offering. 5× faster speeds than Standard processing at twice the price, with no change in intelligence. Fast mode is backward compatible: requests tagged priority will automatically use Fast mode. Using AI efficiently begins with the outcome.

The stakes, cost of error, urgency, and scale determine the right balance of intelligence, speed, reliability, and cost. That balance can change from one step of a workflow to the next. 6 gives businesses much more room to optimize that equation. Luna delivers performance comparable to models that were frontier-class a year ago at roughly 6 cents on the dollar per task, and at nearly nine times the speed.

A coding workflow, for example, might use Sol to resolve uncertainty and define the plan, then use Luna to implement well-specified changes, write and run tests, and evaluate the results. Another workflow may call for a different balance. 6 family expands the range of those choices. Businesses can apply the maximum useful intelligence at every stage while paying the right price for the value it creates.

Delivering that flexibility starts with making every layer behind the models more efficient. Our efficiency edge comes from improving the models, the inference systems that run them, and the agentic harness that connects them to tools and context. 6 models take a more direct path through work. Better routing keeps hardware productive, optimized production software generates tokens more efficiently, and smarter context management helps agents avoid repeating completed work.

Together, these improvements let us complete more useful work with the same compute, reducing the time, tokens, and cost required for each result. 6 Sol is increasingly helping us find and deliver the next round of gains. Within a human-led process, Sol autonomously rewrote and optimized production kernels, designed and ran hundreds of experiments to improve token generation, and monitored training, intervening when problems arose.

The kernel work helped reduce the end-to-end cost of serving the model by 20%, while its experiments increased token-generation efficiency by more than 15%. This work continues, creating a tighter feedback loop: as our models improve and are able to work more autonomously, our ability to improve efficiencies accelerates. 6. Meeting demand for abundant intelligence requires both more compute and more productive compute.

We are building a resilient infrastructure portfolio and matching each workload to the systems best suited to run it. That approach supports both ends of the price-performance curve. At the lower-cost end, the new Luna and Terra prices make high-volume work economical at much greater scale. At the frontier end, Fast mode gives API customers faster access to Sol when response time is important.

Enterprises can move more AI into everyday operations without sacrificing speed on their most consequential work. Large-scale document analysis, customer-interaction classification, and routine implementation can become economical to run broadly, while complex Sol workloads can move faster when the premium is justified. The gains can compound.

In ChatGPT Work and Codex, Free and Go users can access Terra, while Plus, Pro, Business, and Enterprise users can choose Terra and Luna.

Why it matters

On professional work, as measured by Agents’ Last Exam, Luna outperforms Fable 5 at an estimated cost per task nearly 99% lower. In practice, businesses can define the outcome and quality standard they need, then use evaluations to determine where additional intelligence materially improves the result and where faster, lower-cost processing can deliver the same quality.

What to watch

Within a human-led process, more capable models help our technical team find the next generation of improvements, shortening the path to better performance and lower costs. Our strategy remains focused on advancing both capability and efficiency so each generation of intelligence can accomplish more work at a lower cost. 6 Terra and Luna remain available in ChatGPT Work, Codex, and the OpenAI API.