Nvidia CEO Jensen Huang has made a bold declaration following OpenAI’s introduction of GPT-6 Astra, saying “AGI has arrived” and highlighting the extraordinary computing power behind the company’s latest artificial intelligence system.
Huang’s comments, posted on X, have added fresh fuel to the global debate over whether today’s frontier AI models are approaching — or have already reached — artificial general intelligence (AGI). His remarks are particularly significant given Nvidia’s central role in supplying the GPUs and networking infrastructure powering the current AI boom.
Huang Points to Massive AI Compute
Congratulating the OpenAI team, Huang said GPT-6 Astra was “trained on ~100K plus Nvidia Grace Blackwell NVLink72” systems, pointing to the enormous scale of infrastructure required to develop increasingly capable AI models.
He further wrote: “Congratulations @OpenAI team. 400K GPUs coming online next.”
The comments underscore the rapidly expanding computing requirements of frontier AI. Nvidia’s Grace Blackwell platform has become a key part of the infrastructure used by major AI developers to train and operate advanced models, making Huang’s remarks notable not only as praise for OpenAI but also as an indication of the scale of future AI infrastructure deployment.
OpenAI Positions Astra as a Major Leap
OpenAI formally introduced GPT-6 Astra last week, describing the architecture as its most capable and aligned system to date.
The model represents another step in OpenAI’s push to develop AI capable of independently completing increasingly sophisticated digital and analytical tasks. Rather than being restricted to a narrow set of functions, Astra is designed to handle a broader spectrum of computer-based activities.
If systems such as Astra eventually achieve genuine AGI, they could potentially perform intellectual tasks across a wide range of domains at — or beyond — human levels. However, what qualifies as AGI remains widely debated, with no universally accepted definition or benchmark.
Astra Moves Beyond Chat
One of Astra’s most notable capabilities is its ability to use computers to complete complex workflows.
The model is reportedly designed for tasks spanning software engineering, cybersecurity, scientific research and professional work. In OpenAI’s OSWorld 2.0 testing, Astra completed computer-use tasks 47% faster than GPT-5.6 Sol, highlighting its emphasis on both capability and speed.
Astra can reportedly fill online forms, update customer records, conduct web research, analyse information and create websites. It can also generate documents, spreadsheets and presentations, bringing AI closer to functioning as an autonomous digital assistant rather than simply a conversational tool.
Phased Access for Users and Businesses
GPT-6 Astra is being rolled out in phases across ChatGPT Plus, Pro, Business and Enterprise plans. A limited group of organisations is also gaining access through OpenAI’s Daybreak Access programme.
The model is not being offered as a separate standalone consumer product. Developers and businesses can also access Astra through an API, enabling them to integrate its capabilities into their own software and workflows.
AI Race Enters a New Phase
Huang’s “AGI has arrived” statement may ultimately prove to be a bold industry prediction rather than a universally accepted conclusion. Yet the combination of Astra’s expanded computer-use capabilities and the vast computing infrastructure being deployed around it signals how quickly AI is evolving.
The next stage of the AI race may no longer be about models that simply answer questions. It could increasingly centre on systems that can reason, navigate software, make decisions and independently complete real-world digital work.
𝐒𝐭𝐚𝐲 𝐢𝐧𝐟𝐨𝐫𝐦𝐞𝐝 𝐰𝐢𝐭𝐡 𝐨𝐮𝐫 𝐥𝐚𝐭𝐞𝐬𝐭 𝐮𝐩𝐝𝐚𝐭𝐞𝐬 𝐛𝐲 𝐣𝐨𝐢𝐧𝐢𝐧𝐠 𝐭𝐡𝐞 WhatsApp Channel now! 👈📲
𝑭𝒐𝒍𝒍𝒐𝒘 𝑶𝒖𝒓 𝑺𝒐𝒄𝒊𝒂𝒍 𝑴𝒆𝒅𝒊𝒂 𝑷𝒂𝒈𝒆𝐬 👉 Facebook, LinkedIn, Twitter, Instagram