OpenAI ships a new generation in three tiers, with a new mode for subagents. But it's rolling out in phases, after talking with government - and that's exactly the signal.
- OpenAI announced a limited preview of GPT-5.6: Sol (flagship), Terra (balanced) and Luna (fast and cheap), unveiled June 26.
- Prices per million tokens: Sol $5/$30, Terra $2.50/$15, Luna $1/$6; Terra comes close to GPT-5.5 at half the price.
- Rolling out in phases, after coordination with government - control is part of the product.
Three names, three tiers, one strategy. OpenAI announced a limited preview of the GPT-5.6 series - Sol as the flagship, Terra as the balanced everyday model, Luna as the fast and affordable variant. By the company's own numbers, Terra comes close to GPT-5.5 at half the price, while Luna delivers solid capability at the lowest price.
More interesting than the leaderboards is something else. GPT-5.6 brings a new max reasoning mode - gives the model more time to think - and a new ultra mode that goes beyond the single agent and harnesses subagents for complex work. The direction is clear: less 'give me an answer,' more 'split the task and run several agents.'
And here's the crux. OpenAI is shipping Sol with its most robust safety layer yet and starting with a limited preview for trusted partners only - whose participation is shared with government, before a wider release. The company explicitly says it doesn't see this as a long-term norm. Whether it's genuine caution or theater for the public isn't visible from outside. It's doing it because it's the shortest path out the door.
I'll say it straight. The signal isn't the leaderboard - it's the brake. Release your strongest model first to a handful of trusted partners, coordinated with the state, and capability and control come tied in one package. For the top of the industry, this is now how a strong model gets out - quietly, dosed, with government in the room.