OpenAI is ready to expand access to its latest GPT-5.6 model family. In a recent post on X, the company confirmed that GPT-5.6 Sol, Terra, and Luna will become publicly available on Thursday, July 9. If you’ve been itching to try the new models since the limited preview began in late June, you won’t have to wait much longer.
Why the rollout took longer than expected
The announcement marks the end of an unusually restricted rollout. When OpenAI introduced the GPT-5.6 family on June 26, access was limited to about 20 trusted partners after the US government requested time to review the models for a broader release. With preview access now expanding globally ahead of Thursday’s launch, that process appears to be complete.
OpenAI isn’t the only AI company to face additional government scrutiny before a major model release. Earlier this year, Anthropic was forced to suspend access to Claude’s latest Fable and Mythos models in compliance with the US Commerce Department’s export controls. After a similar review, it restored availability on July 1.
What’s new in GPT-5.6
Instead of releasing a single model, OpenAI has split GPT-5.6 into a family of three. Sol is the flagship and most capable model aimed at advanced coding and cybersecurity tasks, Terra offers a more balanced option for everyday workflows, while Luna is the fastest and most affordable variant. The new naming scheme is meant to make it easier for developers to choose between intelligence, speed, and cost.
OpenAI has also said that GPT-5.6 brings improvements to reasoning, coding, and long-running agentic tasks. In addition, Sol includes new Max and Ultra reasoning modes for tackling complex workflows.