OpenAI has launched GPT-6 Astra, its latest and most capable AI model, with major improvements in computer use, coding, research, cybersecurity and complex multi-step tasks. The model is designed to do more than answer questions: it can use computers and browsers, work with software, conduct research and create documents, spreadsheets and presentations.
The launch is also attracting attention because OpenAI says Astra has reached its highest Critical level for cybersecurity capabilities under the company’s Preparedness Framework. This means the model can potentially discover previously unknown security flaws and develop ways to exploit them when given the necessary tools and access.
What Is GPT-6 Astra?
GPT-6 Astra is OpenAI’s newest frontier AI model, built for end-to-end work that requires reasoning, tool use and multiple steps. OpenAI describes it as its most intelligent and aligned model, with state-of-the-art capabilities across computer use, browsing, software engineering, cybersecurity, science and professional work.
Unlike a conventional chatbot that mainly generates text in response to a prompt, Astra is designed to interact with computer environments and carry tasks through to completion. It can fill online forms, update customer records, organise calendars, conduct online research and work inside document and productivity applications.
What Can GPT-6 Astra Do?
OpenAI says Astra can handle a broad range of professional and technical workflows. Its capabilities include:
- Computer and browser use
- Complex reasoning and research
- Software engineering and coding
- Cybersecurity analysis
- Scientific data analysis
- Website creation and frontend testing
- Document, spreadsheet and presentation creation
- Long-running, multi-step tasks using tools
The model can also adapt when users add new requirements or change direction during a task, according to OpenAI’s release notes.
Astra’s Large Context Window
GPT-6 Astra has a 1.05 million-token context window and can generate up to 128,000 output tokens. A large context window allows the model to work with substantially more information within a single task, which can be useful for large codebases, lengthy research material and complex business workflows.
Why Astra’s Cybersecurity Capabilities Are Controversial
The biggest concern surrounding Astra is not simply that it is more capable at coding. OpenAI says the model has crossed its Critical cybersecurity capability threshold, meaning that with appropriate tools and access it can identify previously unknown vulnerabilities and develop new exploitation methods across well-protected systems without a person guiding every step.
This creates a dual-use problem. The same capabilities can potentially help security teams discover and fix vulnerabilities faster, while also increasing the potential impact if such capabilities are misused.
OpenAI says it has therefore strengthened protections around Astra. These measures include stricter isolation, checkpoint encryption, monitoring of full model trajectories and an alignment evaluation process before internal use.
OpenAI Had Already Warned About Astra’s Capabilities
Before the launch, OpenAI published a September 1 assessment explaining that Astra appeared capable of reaching the Critical cybersecurity threshold. The company said it had gathered additional evidence and evaluations before concluding that the threshold had been met.
OpenAI had also previously said in August that internal evaluations of Astra showed significant progress in agentic coding and cybersecurity. At that stage, the company said it could not rule out critical cyber capabilities.
Astra and Computer Use
Computer use is another major focus of the new model. Astra can interact with computer interfaces to complete tasks rather than simply explaining how a user should perform them.
OpenAI reports that in its OSWorld 2.0 latency simulations, Astra achieved a 72.6% score while taking roughly 40 minutes per task, compared with 65.7% and roughly 75 minutes for GPT-5.6 Sol. OpenAI says this represents about a 47% reduction in time per task in that evaluation.
GPT-6 Astra Pricing and Availability
GPT-6 Astra is initially rolling out to a limited set of organisations. OpenAI says it will become available to ChatGPT Plus, Pro, Business and Enterprise users over the following days, alongside access through the OpenAI API, Microsoft Azure and Amazon Bedrock.
For the OpenAI API, the standard listed price is $10 per million input tokens and $50 per million output tokens. The developer documentation also lists a 1.05 million-token context window and a 128,000-token maximum output.
What Astra Means for Indian Users
For users in India, the most important change is the move toward AI systems that can perform complete workflows instead of simply providing information. Developers, researchers, businesses and professionals could use Astra for coding, research, data analysis and document creation.
However, availability may initially depend on the user’s ChatGPT plan, organisation and rollout status. Users should also consider that some advanced capabilities, particularly cybersecurity-related features, may have additional restrictions because of OpenAI’s safety controls.
GPT-6 Astra vs Earlier OpenAI Models
| Area | GPT-6 Astra | Earlier Models |
|---|---|---|
| Computer use | Advanced computer and browser interaction | More limited task execution |
| Coding | State-of-the-art software engineering capabilities claimed by OpenAI | Strong coding capabilities, but less advanced agentic workflows |
| Cybersecurity | Reaches OpenAI’s Critical capability threshold | Lower capability classification |
| Context window | 1.05 million tokens | Varies by model |
| Maximum output | 128,000 tokens | Varies by model |
| Professional work | Designed for multi-step end-to-end workflows | Generally more dependent on user-directed steps |
What Happens Next?
OpenAI’s rollout of Astra is likely to put greater emphasis on both AI agents and AI safety. The company is expanding access while maintaining additional monitoring and controls for potentially harmful behaviour.
The launch also highlights a broader challenge for the AI industry: as models become capable of taking more independent actions, the risks associated with misuse, unexpected behaviour and cybersecurity become more significant. OpenAI’s own safety assessment says Astra required substantially stronger protections because of its capabilities.
Frequently Asked Questions
What is GPT-6 Astra?
GPT-6 Astra is OpenAI’s latest frontier AI model, designed for complex reasoning, coding, computer use, research, cybersecurity and professional workflows.
Why is GPT-6 Astra controversial?
Its cybersecurity capabilities have raised concerns because OpenAI says Astra can potentially find previously unknown vulnerabilities and develop exploitation methods with the right tools and access.
Is GPT-6 Astra available to everyone?
No. OpenAI says the rollout began with a limited group of organisations, with broader availability planned for Plus, Pro, Business and Enterprise users and through the API.
How much does GPT-6 Astra cost?
OpenAI’s API documentation lists standard pricing of $10 per million input tokens and $50 per million output tokens. ChatGPT access depends on the applicable subscription and rollout.
Can Astra use a computer?
Yes. Computer and browser use is one of Astra’s major capabilities. OpenAI says it can perform tasks such as filling forms, updating CRM records, organising calendars, researching online and testing websites.
Bottom Line
GPT-6 Astra represents a significant shift toward AI that can perform multi-step tasks rather than simply generate answers. Its advances in computer use, coding, research and professional work could make it useful for developers and businesses, while its Critical-level cybersecurity capabilities explain why OpenAI has introduced stronger safeguards around the model.
Leave a Reply