OpenAI has launched GPT-6 Astra for computer use, coding, research and complex professional work, with staged ChatGPT access and API pricing of $10 per million input tokens and $50 per million output tokens.
What you need to know
- Available in ChatGPT Work, Codex and the API; enterprise access is off by default at launch
- $10/M input and $50/M output API pricing
- Enterprise administrators can restrict approved websites, desktop applications, uploads, downloads and browsing history
What OpenAI launched
GPT-6 Astra is positioned for computer use, software engineering, research and long, tool-heavy professional workflows. OpenAI's 9 September business update confirms access through ChatGPT Work, Codex and the API. Enterprise administrators must actively enable access under their applicable rate card and agreement because enterprise access is off by default at launch.
The 9 September enterprise control update
OpenAI says administrators can restrict Astra to approved websites and desktop applications, manage uploads and downloads, and control browsing history. ChatGPT Work and Codex can also require confirmation before consequential actions and apply automated review to potentially unsafe or unauthorised tool calls. These are important controls, but they do not remove the need to test role permissions, audit evidence, error recovery and the exact applications an agent can reach.
Price changes the buyer test
Standard API pricing is $10 per million input tokens and $50 per million output tokens, while fast processing is priced at twice the standard rate. That makes a task-level evaluation essential: record tokens, retries, tool calls, completion time and human correction before replacing a cheaper production model.
Who should shortlist it
Astra is most relevant when an agent must navigate software, documents, spreadsheets and websites across a changing multi-step job. Teams buying routine chat, summaries or low-cost drafting should compare the finished-work cost against faster and cheaper models before switching.
What HubAI has not verified
HubAI has not yet completed an independent same-task benchmark. Capability claims in this briefing are attributed to OpenAI; the product remains unscored until Astra is tested against current OpenAI, Anthropic and Google alternatives on identical work.
HUBAI VIEWThe buying decision is total cost per completed workflow—not the launch-day benchmark headline.
Buyer decision signal: UK breakout · 20K+ related searches
What to verify next
1UK availability by subscription tier
2Production API model identifier
3Cache and fast-mode economics
4Enterprise enablement and data controls
5Cost per accepted workflow
Read the evidence
Capabilities, availability and prices can change. HubAI keeps analysis separate from the underlying official material.
01OpenAI: GPT-6 Astra launchOpen source ↗02OpenAI: GPT-6 Astra for professional workOpen source ↗03OpenAI: GPT-6 Astra safety overviewOpen source ↗04Google Trends: United KingdomOpen source ↗
Models
Models
Models
Models