GPT-6 Astra is here.
And the interesting part isn't just the benchmarks.
OpenAI says it reaches:
• 98% on FrontierMath Tier 4 • 99.9% on ARC-AGI 3 • 100% on ExploitBench
But the bigger claim is its focus on real computer use, professional work, science, coding and cybersecurity.
If these capabilities translate into reliable real-world performance, we're moving toward models that can do much more than generate code or answer questions.
They can become actual engineering tools.
The next question is how developers will build around this level of capability.
GPT-6 Astra is here.
We hope it will begin to enable a new generation of entrepreneurship, scientific discovery, and building.
We believe it is the best model in the world for computer use, professional work, science, coding, cybersecurity, and more.
It took us some extra time to ensure that we could meet the safety and alignment standards required for this capability level, but we think you’ll find it worth the wait.
It scores 98% on FrontierMath Tier 4, 99.9% on ARC-AGI 3, and 100% on ExploitBench.
500 followers. 🚀
It might not sound like a huge number, but for me it means a lot.
A few months ago, I started sharing what I’m learning about AI, cloud, DevOps, cybersecurity and developer tools.
Since then, I’ve met a lot of interesting people, learned from the community, and improved a lot along the way.
Really happy to be here and excited to keep building, learning and sharing.
Next stop: 1,000. 🤝
If you’re building in AI, DevOps, cloud or tech in general, I’d love to connect.