By Rajwa Quasim
Nvidia CEO Jensen Huang declared the arrival of artificial general intelligence, or AGI, following OpenAI’s launch of GPT-6 Astra, the company’s latest advanced AI model.
Huang made the comment on X while congratulating the OpenAI team. He said Astra was trained using more than 100,000 Nvidia Grace Blackwell NVLink 72 systems and noted that another 400,000 Nvidia GPUs are expected to come online next. Huang wrote, “AGI has arrived,” referring to Astra.
OpenAI launched GPT-6 Astra last week, its latest and most capable artificial intelligence model. The company’s president, Greg Brockman, described it as a “generational leap.” AGI is generally understood as an AI system capable of performing a wide range of intellectual tasks rather than being limited to a narrow function. In OpenAI’s charter, AGI is described as “autonomous systems that outperform humans at most economically valuable work.”
READ: Nvidia CEO says the US has nothing to fear from China’s open-source AI models (July 23, 2026)
In OpenAI’s OSWorld 2.0 test, Astra completed tasks 47% faster than GPT-5.6 Sol. Astra achieved a 72.6% success rate and took about 40 minutes to complete each task, compared with 75 minutes for GPT-5.6 Sol.
Users can give Astra a goal and have it complete a series of tasks with limited human intervention. It is capable of browsing the internet, operating a computer, using software, writing code, conducting research and performing various other tasks. It is also better equipped to make decisions when instructions are incomplete.
The company said GPT-6 Astra is initially rolling out to a limited set of organizations under OpenAI’s Daybreak Access program. It is available to users on ChatGPT Plus, Pro, Business and Enterprise plans, as well as through the OpenAI API and AWS.
READ: Nvidia may hit $1 trillion through 2027, Huang predicts (March 17, 2026)
Although OpenAI does not explicitly describe Astra as AGI, Brockman said it could be an initial step toward AGI. Meanwhile, OpenAI CEO Sam Altman called AGI “not a super useful term,” adding that it is largely an “irrelevant marketing term.”
Last year, Nvidia and OpenAI announced a letter of intent under which Nvidia would help build at least 10 gigawatts of AI data center capacity for OpenAI, with Nvidia investing up to $100 billion in the project. But in early 2026, Nvidia CEO Jensen Huang clarified that the $100 billion figure was never a commitment. The plan was subsequently stalled.
In February, OpenAI announced a new $30 billion investment from Nvidia as part of a larger funding round. The company said in a release, “We are also expanding our long standing collaboration with NVIDIA, including the use of 3 GW of dedicated inference capacity and 2 GW of training on Vera Rubin systems. This builds on Hopper and Blackwell systems already in operation across Microsoft, OCI, and CoreWeave. Together, this capital and infrastructure expansion strengthens our ability to train and deploy frontier models at global scale.”