WASHINGTON (dpa-AFX) - OpenAI has introduced GPT-6 Astra, its latest frontier artificial intelligence model, describing it as its most intelligent and aligned model to date.
Astra is designed to deliver major improvements in computer use, software engineering, cybersecurity, scientific research and professional work.
The model achieved a 98 percent score on FrontierMath Tier 4, 99.9 percent on ARC-AGI-3 and 100 percent on ExploitBench. On Terminal-Bench Science 0.1, Astra scored 64.6 percent, compared with 52.6 percent for Claude Fable 5.1, while costing about 31 percent less in the tested configuration. On Terminal-Bench 4.0, it reached 57.9 percent, compared with 37.3 percent for GPT-5.6 Sol.
Astra also showed stronger performance on computer-based professional tasks. It scored 59.3 percent on Agents' Last Exam, ahead of Claude Opus 5 at 55.5 percent and GPT-5.6 Sol at 53.6 percent.
OpenAI says the model can handle tasks such as creating websites, analysing data, producing presentations and spreadsheets, testing software and completing online forms.
The company also highlighted Astra's scientific and cybersecurity capabilities, including a 96 percent score on GPQA Diamond. At the same time, OpenAI says the launch version includes safeguards against advanced offensive cybersecurity tasks.
Astra is initially rolling out to a limited group of organisations before becoming available to ChatGPT Plus, Pro, Business and Enterprise users.
It will also be offered through the OpenAI API, Microsoft Azure and Amazon Bedrock. API pricing starts at $10 per million input tokens and $50 per million output tokens.
Copyright(c) 2026 RTTNews.com. All Rights Reserved
Copyright RTT News/dpa-AFX
© 2026 AFX News
