OpenAI is preparing to launch a new model series: What to know

OpenAI is all set to launch a new model series named “Astra” to enhance multi-task agent task collaboration, a significant move that these models will enable multiple agents to work together to solve complex problems in key areas such as mathematics, project management Astra following previous releases like Sol, Terra and Luna.

At present details regarding the release timeline whether it is GPT-6 remain unclear while testing is underway, and the models could be the first to adhere to the new CFT framework proposed by the Trump administration, which require AI models to be submitted to pre-release clearance prior to public release. This development could weigh on investment portfolio and crypto markets as supervisory pressure intensifies.

According to reports, OpenAI founder Sam Altman presented “Astra” to policymakers and regulators in Washington this week. The series is promoted as being capable of enabling multi-agent coordination to collaborate over the long term on advanced tasks such as mathematical problems. Astra will be a new class of models after OpenAI’s Sol, Terra, and Luna.

The models are already in the testing phase, and they are intended to undergo through the Trump administration's new framework requiring AI models to be submitted to the federal government before official release. Furthermore, OpenAI plans to release a report about how its AI can solve complex mathematical problems illustrating the capabilities of its most advanced technology.