OpenAI is reportedly preparing a new family of artificial intelligence models that may use the name Astra. The reported system is said to improve on handling longer-running tasks, pointing to a push beyond shorter prompt-and-response interactions toward more sustained work.
According to the report, OpenAI CEO Sam Altman previewed Astra in Washington, DC. The demonstration was aimed at policymakers, suggesting the company is continuing to discuss its technology not only with customers and developers, but also with officials involved in shaping AI rules and oversight.
If Astra launches as described, it could mark another step in the race to build AI tools that can manage more complex assignments over time. That kind of capability has become a major focus across the industry, as companies try to make models more useful for real-world workflows that require persistence, planning and follow-through.
OpenAI has not publicly detailed the model family in the available report, and the name Astra was described as tentative. Even so, the Washington preview signals that OpenAI may be getting closer to introducing another major product update while keeping regulators and lawmakers in the loop.
