OpenAI introduced GPT-6 Astra, its most intelligent and aligned model, now available in ChatGPT Work, Codex, and the API. Astra excels in computer use, professional work, software engineering, cybersecurity, and science, enabling teams to handle demanding tasks with speed and accuracy.

Unlike other AI systems, Astra works directly within existing applications, even those without APIs, eliminating the need for data preparation or workflow redesign. Early adopters are using it to optimize GPUs, spot financial discrepancies, and produce branded decks. Internally, OpenAI teams used Astra to convert three hours of footage into a developer video, achieving over 550k views in four days.

Astra also resolved a memory bottleneck in Codex, reducing latency by 25x while increasing peak memory by 30%. It follows company voice and design standards better, delivering more useful work per dollar. With fewer tokens and retries, it lowers costs, leading OpenAI to dominate cost-efficiency benchmarks like Terminal Bench 4.0.

Pricing is set at $10 per million input tokens and $50 per million output tokens. Safety is enhanced: Astra shows 89% fewer unintended outcomes than GPT-5.6 Sol and 74.7% fewer than Claude Fable 5.1. New enterprise controls allow restricted access, approval workflows, and automated review of tool calls, ensuring secure deployment.