ChatGPT's GPT-6 Pro Can Now Run Several Agents in Parallel

OpenAI said on October 16, 2026 that GPT-6 Pro in ChatGPT can now divide a single task among several agents working at the same time. Users can follow each agent’s progress and open finished subtasks to review the work.

What changed?

According to OpenAI’s announcement on X, the Pro model got “faster responses and improved use of plugins and tools,” and now lets multiple agents tackle different parts of a task in parallel. The feature lives in ChatGPT’s regular chat experience. OpenAI’s posts have referred to the model as both “GPT-6 Pro” and “GPT-6 Astra Pro”; ChatGPT’s pricing page lists Pro reasoning with GPT-6 Astra as a Pro-plan feature.

The rollout targets eligible Pro subscribers. It comes nine days after OpenAI made GPT-6 and its “Intelligent UI” interactive answers available to all ChatGPT users on October 7.

What has OpenAI not said?

OpenAI has not published latency or accuracy figures, the maximum number of agents per task, or how parallel runs count against usage limits. Early user reaction compiled by Digg was mixed: some praised the speed, while others complained about usage limits and tier changes.

Why it matters: Parallel agents move ChatGPT closer to delegating whole projects rather than answering single prompts, which is where competition with Claude, Gemini and Grok is heading. For teams weighing plans, see our ChatGPT vs Gemini vs Claude for business comparison.