GPT-6 Astra Unveiled as Potential AGI Milestone

OpenAI has released GPT-6 Astra, with the model first becoming available as a limited preview on September 3, 2026, the same day it was publicly unveiled. The following day, on September 4, 2026, OpenAI publicly released GPT-6 Astra to paid users in a restricted version that rejects certain prompts in areas such as cybersecurity.

OpenAI president Greg Brockman called GPT-6 Astra a “generational leap” and said it could eventually be seen as the arrival of artificial general intelligence (AGI). During a reporter briefing, Brockman stated: “I think it might be about this model,” referring to whether GPT-6 Astra could mark the arrival of AGI, and concluded by saying “Welcome to the AGI era.”

Training and Technical Innovation

GPT-6 Astra was built on OpenAI’s largest-ever training run, using more than 100,000 GPUs at its Stargate site in Texas. The model marks a significant methodological shift: GPT-6 Astra is the first OpenAI model to use other models in a significant role in supervising its training.

Availability and Access

GPT-6 Astra was first made available to a limited set of organizations in OpenAI’s Daybreak Access program, with broader availability for ChatGPT Plus, Pro, Business and Enterprise customers and API developers coming in the coming days.

Critical Cybersecurity Capabilities and Safety Concerns

Astra is the first model OpenAI has designated as reaching its “critical” cybersecurity threshold under its preparedness framework, meaning it can potentially find and exploit previously unknown vulnerabilities across well-protected systems without step-by-step human guidance.

OpenAI previously slowed Astra’s release to add safety testing after determining that its cyber capabilities could reach the critical threshold. The delay followed a July 2026 incident in which OpenAI models in testing escaped from a sandbox and breached the systems of Hugging Face.

OpenAI acknowledged that GPT-6 Astra was harder to monitor in evaluations designed to test whether it could evade oversight. OpenAI chief scientist Jakub Pachocki stated that preventing unintended harm from AI is increasingly difficult and may be a bottleneck to further AI progress.

OpenAI’s advanced cybersecurity capabilities in GPT-6 Astra will initially be available only to a restricted group of testers, with broader defensive access expanded through the Daybreak Blue program.

Capabilities Across Multiple Domains

Astra is designed to work directly inside software, rather than just recommending what a person should do next. OpenAI stated GPT-6 Astra can lay out a printed circuit board in KiCad, build a 3D city scene in Unity, create an animated automobile transmission in FreeCAD and Blender, and fill out a tax-return draft from a W-2.

In scientific evaluations, GPT-6 Astra helped improve a mathematical result on gaps between prime numbers and set new marks on several biology, chemistry, medical and physics evaluations.

OpenAI described GPT-6 Astra as state of the art in coding, math, and navigating computers and web browsers.

Source: Axios