OpenAI releases GPT-6 Astra, its latest AI model
AI firm OpenAI has launched its newest model, GPT-6 Astra, designed to perform tasks more autonomously using computers. The model reportedly achieves high scores on reasoning and professional task benchmarks.
San Francisco – OpenAI has launched GPT-6 Astra, its latest frontier model, aiming to advance AI systems capable of using computers and executing tasks more autonomously. The company stated that Astra is its most capable model to date.
The model supports reasoning capabilities across various levels and features a context window of 1.05 million tokens with a maximum output of 128,000 tokens. OpenAI claims Astra can autonomously navigate websites, interact with software, work with documents and spreadsheets, write and test code, and handle other multi-step workflows. It is positioned as a significant upgrade for AI interface capabilities.
According to OpenAI, Astra achieved high scores on benchmarks such as FrontierMath Tier 4 (98%) and ARC-AGI-3 (99.9%), demonstrating advanced capabilities in mathematics and reasoning. The company asserts that the model handles demanding professional work with exceptional speed, accuracy, and judgment.
Initially, GPT-6 Astra will be available to a select group of organizations before rolling out to all ChatGPT Plus, Pro, Business, and Enterprise users. The model's knowledge cutoff is April 30, 2026, and it supports tools including web search, file search, computer use, and image generation.
The release emphasizes OpenAI's focus on agentic AI, positioning Astra to compete with models from companies like Anthropic and Google. This move signifies a push towards AI agents that can execute complex tasks independently, moving beyond simple prompt-based interactions.