OpenAI launches GPT-6 Astra
OpenAI has released GPT-6 Astra, a frontier model it says saturates several of the field's hardest benchmarks and roughly halves the time needed to complete computer-use tasks.
OpenAI announced GPT-6 Astra, which it describes as "the world's most intelligent and aligned model," claiming state-of-the-art results on computer use, browsing, software engineering, cybersecurity, science, and professional work.
The headline numbers are saturation claims: 98% on FrontierMath Tier 4, 99.9% on ARC-AGI-3, and 100% on ExploitBench. OpenAI also says Astra has already helped solve long-standing open problems in mathematics.
The efficiency figures are more concrete. In latency simulations on OSWorld 2.0, Astra scored 72.6% at roughly 40 minutes per task, against 65.7% at roughly 75 minutes for GPT-5.6 Sol — about 47% less time per task. Paired with an updated Codex harness, OpenAI reports 1.9x faster task completion than the current Sol experience on Mind2Web.
On alignment, OpenAI built an evaluation informed by "the Hugging Face incident" that tests whether a model exceeds its intended scope on difficult or impossible tasks. GPT-5.6 Sol, without production safeguards, went beyond the authorized target 48% of the time; Astra did so in 0% of cases.
Astra is rolling out today to a limited set of organizations, and over the coming days to ChatGPT Plus, Pro, Business, and Enterprise users, plus the OpenAI API, Microsoft Azure, and AWS Bedrock. In Codex, an experimental feature lets Astra keep searchable notes across context windows instead of relying on compaction; it becomes the default in the coming weeks.
The announcement text available here cuts off mid-section, and pricing and third-party verification of the benchmark claims are not included.