US media: OpenAI is preparing to release a new model series Astra focuses on multi-intelligent body length cycle task collaboration

On August 1, TheInformation reported that according to three people familiar with the matter, OpenAI is preparing to release a new model series, tentatively named "Astra", which will improve its ability to complete long-cycle tasks. OpenAI founder Altman has demonstrated Astra to policymakers and regulators in Washington this week. OpenAI is said to promote its ability to have multiple agents work together over time to solve particularly difficult problems, which can be applied to projects or advanced mathematical problems. Astra will become the new class of OpenAI model after Sol, Terra and Luna. It is unclear when OpenAI plans to release the model. OpenAI has also not yet decided whether Astra will be labeled as GPT6 or as an add-on model to the GPT5 family (e.g. GPT 5.7). According to one of the people familiar with the matter, although the models are already in testing, they are expected to be the first to be submitted under the Trump administration's planned new framework, which requires AI models to be submitted to the federal government before being released to the public. The government aims to finalize the framework before the end of this week's self-imposed deadline. In the short term, OpenAI is said to also plan to release a report explaining how it solved 10 previously unsolved mathematical problems to demonstrate its state-of-the-art AI capabilities. (Jin Shi)