OpenAI is preparing to release a new model series, Astra, focusing on multi-agent long-horizon task coordination.

August 1st - According to three sources familiar with the matter, OpenAI is preparing to release a new model series, tentatively named "Astra," which will enhance its ability to complete long-term tasks. OpenAI founder Altman presented Astra to policymakers and regulators in Washington D.C. this week.

OpenAI is said to promote its ability to enable multiple agents to collaborate on long-term solutions to highly challenging problems, applicable to projects or advanced mathematical issues. Astra will be a new class of models from OpenAI following Sol, Terra, and Luna. It is currently unclear when OpenAI plans to release this model. OpenAI has also not decided whether Astra will be labeled as GPT-6 or introduced as an additional model to the GPT-5 series (such as GPT-5.7).

According to one of the sources, although these models have been in testing, they are expected to be among the first models submitted under a new framework planned by the Trump administration, which requires AI models to be submitted to the federal government before public release. The self-imposed deadline set by the government is this weekend, with the goal of finalizing the framework by then. It is said that OpenAI also plans to release a report in the near future detailing how it has solved 10 previously unsolved math problems to showcase the capabilities of its cutting-edge AI. (The Information)