According to CoinMeta, as reported by The Information, OpenAI is preparing to release a new model series called “Astra”, which will see improvements in its capabilities for long-term tasks. The founder of OpenAI, Altman, demonstrated Astra to policymakers in Washington this week, highlighting its ability to enable multiple agents to work together over the long term to solve complex problems. Astra will become a new category of models following SOL, Terra, and Luna. It is not yet clear when OpenAI plans to release this model, whether it will be labeled as GPT-6, or if it will be an additional model in the GPT-5 series (such as GPT-5.7). Sources familiar with the matter have revealed that these models are already in testing and are expected to be among the first to be submitted under the new framework planned by the Trump administration, with the government setting a deadline for this weekend. In addition, OpenAI also plans to release a report explaining how these models can solve 10 unsolved mathematical problems, to demonstrate their AI capabilities.