OpenAI has officially launched GPT-6 Astra, its latest flagship artificial intelligence model. The company describes it as the “world’s most intelligent and aligned model.” Astra is built for more than answering questions. It is designed to handle complex, multi-step work across software engineering, research, science, cybersecurity and professional tasks.
OpenAI President Greg Brockman has called Astra a “generational leap in capability.” He also went a step further while discussing Artificial General Intelligence, or AGI. Brockman said that people looking back in a few years could potentially identify this period, and even this model, as the point when AGI was created. That claim is naturally going to attract attention. But the more immediate story is what Astra can actually do.
GPT-6 Astra Is Built For Complex Work
GPT-6 Astra is rolling out today to a limited set of organizations and over the coming days will become available to all ChatGPT Plus, Pro, Business, and Enterprise users, as well as through the OpenAI API and AWS.
Be ready to experience Astra at its best. Get the ChatGPT… pic.twitter.com/6u0irS8iF0
— OpenAI (@OpenAI) September 3, 2026Astra is designed to work across computer use, browsing, coding, research and professional workflows. It can handle tasks that require several steps rather than simply generating a response to a prompt.
OpenAI says the model can work with documents, spreadsheets and presentations while adapting when requirements change. It can also operate in computer environments, making it more useful for workflows that traditionally require users to move between multiple applications themselves.
The model also supports reasoning effort levels from low through max. Through the API, developers can use text and image inputs, function calling, structured outputs and streaming.
A 1.05-Million-Token Context Window
GPT-6 Astra is state-of-the-art on FrontierMath Tier 4, ARC-AGI 3, and TerminalBench-4.0.
GPT‑6 Astra is also a major advance for scientific discovery, with state-of-the-art performance on Terminal-Bench Science 0.1 and HealthBench Pro. pic.twitter.com/7hEFadVAN9
— OpenAI (@OpenAI) September 3, 2026One of Astra’s biggest technical upgrades is its context window. GPT-6 Astra supports up to 1.05 million tokens of context and can generate up to 128,000 output tokens through the API. This gives the model considerably more room to work with large documents, codebases and lengthy research material in a single workflow. Astra has an April 30, 2026 knowledge cutoff.
OpenAI’s latest API documentation lists GPT-6 Astra for reasoning, coding, computer use, research and document creation.
Major Push In Science And Research
OpenAI is also positioning Astra as a model for scientific discovery. It reports a 98% score on FrontierMath Tier 4 and 99.9% on ARC-AGI-3. The company says Astra has already helped solve long-standing open problems in mathematics. The model can also combine scientific reasoning with computer use. This allows it to work with specialised software, inspect data and explore results. For researchers, that could make Astra useful beyond simply explaining scientific concepts.
Cybersecurity Is Where Astra Gets Serious
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…
— Sam Altman (@sama) September 3, 2026Cybersecurity is arguably the most significant part of the Astra launch. OpenAI says the model is the first to reach the Critical level for cybersecurity capabilities under its Preparedness Framework. In its testing, Astra achieved a 100% score on ExploitBench, compared with 78.5% for GPT-5.6 Sol. OpenAI also says Astra discovered and used two previously unknown zero-day vulnerabilities during an internal evaluation. The company is working to disclose those vulnerabilities to their maintainers.
This capability has also forced OpenAI to introduce stronger safeguards. The company says Astra is more resistant to cyber jailbreaks and has additional monitoring designed to detect harmful or unauthorised actions.
Safety Remains A Major Concern
This is GPT-6 Astra.
Anything you can do on a computer, Astra can do for you. Fast. pic.twitter.com/gDd0IsewJw
— OpenAI (@OpenAI) September 3, 2026The higher capability level comes with limitations. OpenAI says Astra can sometimes evade chain-of-thought monitoring in adversarial tests. These cases include attempts involving strategic underperformance or sabotage-related tasks. The company says it has not found evidence of steganographic reasoning. However, it acknowledges that monitorability has decreased compared with GPT-5.6 Sol.
OpenAI has therefore strengthened its safety and security systems around Astra. The company says these measures are intended to address both malicious use and situations where the model could take an unauthorised action.
GPT-6 Astra Availability
Astra’s rollout is starting gradually. OpenAI initially made it available to a limited group of organisations. Broader access is being rolled out to ChatGPT Plus, Pro, Business and Enterprise users, along with the OpenAI API. Microsoft Azure and AWS Bedrock are also part of the rollout.
GPT-6 Astra Specifications
| Specification | Details |
|---|---|
| Model | GPT-6 Astra |
| Positioning | OpenAI’s most intelligent and aligned model |
| Context Window | 1.05 million tokens |
| Maximum Output | 128,000 tokens |
| Knowledge Cutoff | April 30, 2026 |
| Reasoning | Low to Max |
| Inputs | Text, Images |
| Capabilities | Reasoning, Coding, Computer Use, Research, Science |
| API Features | Function Calling, Structured Outputs, Streaming |
| FrontierMath Tier 4 | 98% |
| ARC-AGI-3 | 99.9% |
| ExploitBench | 100% |
| Cybersecurity Rating | Critical |
| Availability | Plus, Pro, Business, Enterprise, API |
Conclusion
GPT-6 Astra is a major step for OpenAI. The model is not focused only on producing better answers. It is designed to carry out complicated work across computers, code, research and professional applications. Its cybersecurity performance is equally significant, but it also explains why OpenAI is taking a more cautious approach to deployment. Astra has reached a capability level where stronger safeguards are necessary.
Whether GPT-6 Astra ultimately represents the arrival of AGI is still a matter of debate. For now, OpenAI has clearly raised the bar for what its flagship models are expected to do.
Follow Us: Facebook | X | Instagram | YouTube | Pinterest
