OpenAI Upgrades GPT-6 Astra Pro in ChatGPT with Faster Responses and Parallel Agents

OpenAI has announced an upgrade to GPT-6 Astra Pro inside ChatGPT, adding faster responses, better use of plugins and tools, and a new ability to run multiple AI agents in parallel on a single task. The announcement was posted by the verified OpenAI account on 16 October 2026.

The update is available to ChatGPT Pro users in the Chat tab, according to OpenAI’s own release notes. Access to GPT-6 Astra varies by plan and workspace permissions, so not every ChatGPT user will see the new features straight away.

What GPT-6 Astra Pro Actually Does

OpenAI describes GPT-6 Astra as a model built for coding, research, computer use, and complex multi-step work. GPT-6 Pro — the tier most users interact with — runs on GPT-6 Astra under the bonnet.

The parallel-agent feature is the headline addition. Rather than working through a long task in sequence, GPT-6 Astra Pro can now split the job across several agents, each handling a different part at the same time. Users can watch each agent’s progress inside the conversation as it happens, then open the completed results to review what each one did.

It’s the kind of workflow that could matter for anyone managing a research project, a software build, or a lengthy document — tasks where waiting for one step to finish before starting the next costs real time.

Speed and Tool Use

OpenAI says responses are now faster, and that the model makes better use of plugins and connected tools. The company hasn’t published a numerical benchmark or percentage improvement — so there’s no independent way to verify how much faster “faster” actually is. No third-party performance tests were identified at the time of publication.

The tool-use improvement matters because GPT-6 Astra Pro can connect to external services — search, code execution, data retrieval — to complete tasks. Better tool handling, if the claim holds up in practice, could mean fewer errors when the model hands off work between those services.

What OpenAI Hasn’t Said

The announcement is light on detail in a few areas that will matter to organisations thinking about deploying the feature.

OpenAI hasn’t said how many agents can run in parallel, or what happens when two agents reach conflicting conclusions. There’s no information on how the system logs agent decisions for audit purposes — a question that’s not academic for businesses or public bodies that need to show their working.

Data protection is another open question. When multiple agents access connected tools and external services, the scope of what data those agents can see — and where it goes — widens. Organisations that use ChatGPT under enterprise agreements will need to check whether the parallel-agent feature falls within their existing permissions and data-handling arrangements.

The Bigger Picture

OpenAI is one of several AI companies pushing hard on what the industry calls “agentic” AI — systems that don’t just answer questions but carry out multi-step tasks with some degree of autonomy. Google, Anthropic, and others are working along similar lines.

The parallel-agent approach is a direct response to a real frustration with current AI tools: complex tasks take time, and users have to wait. Splitting work across agents is a practical attempt to cut that wait.

But speed alone doesn’t settle the deeper questions. Who checks the agents’ work? What happens when one agent’s output feeds incorrectly into another’s? OpenAI’s announcement doesn’t address those points, and they’re the ones that will determine how far organisations are willing to trust the system with consequential tasks.

Sam Altman, OpenAI’s chief executive, has said previously that agentic AI represents the next major step for the company — moving ChatGPT from a tool that answers questions to one that completes entire projects. This update is a step in that direction, though the gap between a demo and a reliable production tool remains wide for many use cases.

What This Means for Kent Residents

Kent residents who subscribe to ChatGPT Pro may be able to use the upgraded GPT-6 Astra Pro features in the Chat tab, depending on their plan and rollout status. Local businesses, schools, and councils considering the tool should check whether their existing data-protection policies cover the use of parallel agents and connected plugins before putting any sensitive information near it — patient records, council data, and commercially confidential material are all in that category. No UK pricing changes were announced alongside the upgrade.