Nvidia CEO Jensen Huang Claims GPT-6 Astra Is AGI

The world may be on the verge of change. After years of debate regarding AGI (Artificial General Intelligence), Nvidia CEO Jensen Huang believes that OpenAI's new AI model, GPT-6 Astra, may have finally reached this stage—potentially representing one of the most significant shifts in the AI industry.

In a post on X, Huang congratulated OpenAI on developing Astra just four years after the AI startup launched ChatGPT, declaring: "AGI has arrived." He added that Astra was "trained on more than 100,000 Nvidia Grace Blackwell NVLink72 systems." The Nvidia CEO revealed that the company planned to bring more than 400,000 additional GPUs online soon.

Is GPT-6 Astra AGI?

AGI, or Artificial General Intelligence, is a term used broadly to describe an AI model capable of thinking and reasoning like a human, or matching human capabilities across various tasks. Experts note that AGI has the potential to surpass human knowledge. As AI models have improved, achieving AGI has become a goal for several AI labs, including OpenAI and Anthropic.

However, AGI remains a subject of debate. OpenAI defines AGI as "highly autonomous systems that outperform humans at most economically valuable work."

Jensen Huang's comments suggest that GPT-6 Astra could represent the most significant recent breakthrough in cutting-edge AI models. OpenAI President Greg Brockman agreed with Nvidia's top executive. "We are entering the era of AGI (whether one considers this model, the previous one, or the next) and we couldn't do it without close partners," he replied. Nvidia remains one of OpenAI's key partners. According to OpenAI, the majority of its training and inference infrastructure continues to run on Nvidia GPUs.

Previously, OpenAI President Greg Brockman stated: "It is not far-fetched to think that we are now—with Astra—in the era of AGI." "I think that if we fast-forward a couple of years and look back to ask, 'When was AGI actually created?', the answer will point to this moment and, possibly, to this model," Brockman explained.

However, not everyone agrees that GPT-6 Astra marks the beginning of the AGI (Artificial General Intelligence) era. AI critic Gary Marcus wrote on X: "It is unfortunate to see Jensen claim that AGI has arrived, without evidence or definitions." Marcus maintained: "By conventional definitions, Astra does not yet reach that level. Declaring victory without a definition simply muddies the waters."

How good is GPT-6 Astra?

OpenAI describes Astra as its most powerful AI model to date. According to the company, Astra can perform tasks related to computer use, software engineering, cybersecurity, science, and other professional activities. In performance tests, the AI model has also outperformed cutting-edge rival models, including Anthropic's Mythos 5.1.

In another post, Thibault Sottiaux, head of Codex at OpenAI, stated that Astra is so good that the company plans to launch new products six months ahead of the original schedule.

At the same time, OpenAI claims to have implemented additional safety measures around Astra to help mitigate cybersecurity risks. The company states that Astra is its "most aligned model to date." In evaluation tests based on Hugging Face's security benchmark, OpenAI found that GPT-6 Astra exceeded authorized objective limits in 0% of cases, compared to the 48% recorded by GPT-5.6 Sol without production safety measures.

GPT-6 Astra is available to users on ChatGPT Plus, Pro, Business, and Enterprise plans. The model can also be used via the API.