OpenAI has begun a limited preview of its GPT-5.6 model series for a small group of trusted partners. The rollout follows a request from the US administration. Broader availability is planned in the coming weeks.
OpenAI started previewing GPT-5.6 to a limited group of trusted partners whose participation has been shared with the government. The series includes three variants: Sol, described as the company's strongest model yet; Terra, for everyday use with performance similar to GPT-5.5 at half the cost; and Luna, the lowest-cost option.
The company gave the US government an early preview of the models. It cited a recent AI cybersecurity order signed by President Trump that encourages voluntary government review of powerful models 30 days before public release. OpenAI stated it does not believe this kind of government access process should become the long-term default.
Sol introduces a max reasoning effort mode and is positioned as the most capable for cybersecurity tasks such as finding and fixing vulnerabilities. All variants include strengthened safeguards against jailbreaks and prohibited cyber assistance after OpenAI spent 700,000 GPU hours testing for weaknesses.
Pricing starts at $5 per million input tokens and $30 per million output tokens for Sol, with lower rates for Terra and Luna. The models will reach wider release in the coming weeks.
