OpenAI Launches GPT-6 Astra and Redefines the AGI Concept

OpenAI has officially launched GPT-6 Astra, an artificial intelligence model that the company claims sets a new standard for performance in software engineering, science, and cybersecurity. This strategic move is designed to reclaim technological leadership in the sector, particularly in the face of direct competition from Anthropic. The Astra model represents more than an incremental update; it incorporates significant advancements in pre-training, reinforcement learning, and alignment, allowing the tool to execute complex tasks with greater autonomy and precision.

A key technical highlight of Astra is its ability to interact directly with computers and web browsers. In tests conducted within the OSWorld 2.0 environment, the model proved to be 47% faster per task than its predecessor, the GPT-5.6 Sol. Furthermore, in the Mind2Web benchmark, Astra achieved a task completion speed 1.9 times faster. For developers, the model excels in performance on real-world codebases, reaching 74.1% efficiency in the DeepSWE v1.1 benchmark, outperforming the 72.7% previously recorded. This technological evolution mirrors broader industry trends, such as the shift toward more secure and autonomous firmware and data management systems.

OpenAI leadership describes the launch as a generational leap that touches upon the concept of Artificial General Intelligence (AGI). While the definition of AGI remains a debated topic, the company suggests that this historical moment will be remembered as the turning point where machines began to surpass human capabilities across a range of cognitive tasks. The global tech landscape is currently witnessing a surge in investment, with major players and startups alike racing to define the future of AI infrastructure, cybersecurity, and digital sovereignty.

The new model also introduces experimental features in Codex, allowing the system to retain information across different context windows. This enables Astra to recover requirements, corrections, and test results during long-term projects, optimizing professional workflows. With this release, OpenAI reaffirms its commitment to innovation, treating AGI not just as a technical goal, but as a core concept of its corporate mission, strengthening its position ahead of planned public offering. As the company continues to refine its models, the focus remains on creating systems that are not only faster and more autonomous but also capable of handling the intricate demands of modern scientific research and complex software architecture. The era of AGI, while still evolving, has clearly entered a new, more tangible phase with the arrival of Astra.