Back to timeline

OpenAI Launches ChatGPT Work

OpenAI launches ChatGPT Work, extending agentic execution across apps, files, and finished workplace deliverables.

Product Launch

What Happened

On July 9, 2026, OpenAI introduced ChatGPT Work, an agent for completing multi-step projects across applications and files. Web and mobile rollout began with Pro, Enterprise, and Edu, with Plus and Business scheduled to follow over the next few days.

Why It Matters

The launch brought technology developed for coding agents into general knowledge work. It made producing and revising deliverables across connected tools a central ChatGPT workflow, rather than limiting assistance to instructions or draft text.

Technical Details

At launch, Work used GPT-5.6 and incorporated Codex technology. OpenAI described tasks that could run for hours, divide projects into smaller steps, and produce spreadsheets, presentations, documents, and web applications. Users could follow progress, provide new direction, and approve important actions. Scheduled Tasks supported continuing recurring workflows while users were away. These features describe product scope, not a guarantee that every long-running task succeeds.