OpenAI will publicly release all three GPT-5.6 variants — Sol, Terra, and Luna — this Thursday, July 9, expanding from a limited partner preview that began in late June. The staggered rollout followed a Trump administration executive order requiring companies to submit their most powerful models for government review 30 days before public deployment. OpenAI complied but objected to the process becoming a long-term standard.

The Department of Commerce's Center for AI Standards and Innovation conducted additional testing after the initial review, with OpenAI sending technical experts to Washington to address questions directly. That engagement accelerated the wider release well within the original 30-day window. Sol leads the lineup as OpenAI's strongest model yet at $5 per million input tokens and $30 per million output. Terra targets everyday use, matching GPT-5.5 performance at half the cost — $2.50 input and $15 output. Luna serves as the budget tier at $1 input and $6 output.

Anthropic faced similar restrictions, temporarily blocking foreign nationals from accessing its Mythos and Fable models before receiving clearance to redeploy Mythos 5 and eventually Fable 5. The episode establishes a precedent where frontier model releases now involve mandatory government checkpointing, adding friction and uncertainty to deployment timelines. For users, the tiered pricing makes advanced capabilities more accessible, but the regulatory overlay means future launches could face unpredictable delays regardless of technical readiness.