OpenAI has confirmed that GPT-5.6 will launch this Thursday, July 9, ending a weeks-long hold prompted by U.S. government review. The company unveiled the model family in late June but restricted early access to a small group of trusted partners at federal request — a departure from its typical release cadence. Now OpenAI says it is "expanding preview access globally," though whether Thursday's drop reaches users outside the United States remains unclear.

The lineup introduces a new naming convention: Sol as the flagship model, Terra as the balanced tier for everyday work, and Luna as the fast, affordable option. OpenAI explains that the version number denotes generation while Sol, Terra, and Luna represent durable capability tiers that can advance independently, giving developers clearer trade-offs across intelligence, speed, and cost.

Sam Altman stated last month that a global rollout was the goal, and Tuesday's social-media update suggests progress toward that target. The government-mandated pause marks the first time a major U.S. AI lab has publicly delayed a flagship release for federal scrutiny, setting a precedent for how future model deployments may be governed.

For consumers and developers, the tiered structure means more predictable pricing and performance choices, but the regulatory friction signals that frontier-model releases will increasingly involve compliance checkpoints — not just engineering milestones.