OpenAI is rolling out ChatGPT 5.6, which rolls out to a limited group of partners globally today and will be available to everyone tomorrow. The company wants to offer us a more powerful and versatile model, but also a safer one, since the entire GPT‑5.6 family comes with new layers of security to make the most of the new capabilities responsibly.
Three models that adapt to every need
With GPT‑5.6, OpenAI introduced three versions: Sol, Terra and Luna. The Sol version is the most powerful and is designed for complex work that requires deep reasoning, such as programming, biological analysis and cybersecurity. Terra offers the best balance between performance and cost, while Luna stands out for its speed.
GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday.
We’re expanding preview access globally now. pic.twitter.com/Uk5HcfSc2e
— OpenAI (@OpenAI) July 8, 2026
After Antrophic temporarily withdrew its models and then reinstated them, OpenAI followed a similar launch strategy with GPT 5.6. For now, only a few trusted partners have access to the system. With OpenAI’s announcement, today access for these partners will expand internationally, and, tomorrow, it will be available to the public.
At first, the models will be available through the API and Codex, with plans to come to ChatGPT and the other platforms very soon. At that point, the rollout will be complete and we’ll get the update we’ve been waiting months for. More powerful tools and safer ones in our daily workflow.
