ChatGPT Work vs Dots, When to Use Which?

The difference between ChatGPT Work and Dots is who controls the timing. Work is an agent that carries a task you start through to a finished document, spreadsheet, presentation, or report. A dot is an always-on agent with its own computer and memory that keeps work moving even when you don't have ChatGPT open. In short, Work is for tasks, a dot is for ongoing responsibilities.

If you're new to Work, I covered how it differs from regular Chat in ChatGPT Now Has Chat and Work, So Where Does Office Work Go?.

The differences in one table

Sources: Work and Codex help page, Dots help page, and the ChatGPT pricing page.

Work, for tasks that end in a finished deliverable

According to OpenAI, Work is designed for longer, multi-step work: researching a topic, analyzing information, or creating a document, spreadsheet, presentation, report, or Site. While it works, you can follow its progress, answer its questions, change direction, and approve important actions.

Work can also run once, repeat on a schedule, or be triggered by an event in a connected app, such as a new Gmail message, a new message in a Slack channel, or pull request activity on GitHub.

Tasks that suit Work:

• Building a monthly sales report from three spreadsheets.
• Researching competitors and turning it into a presentation.
• Preparing a quotation from a client brief.
• Producing an automatic weekly summary every Monday morning.

A dot, for work that needs to be kept going

A dot is powered by GPT-6 Astra, has its own cloud computer, and brings context from your ChatGPT memory. OpenAI says a dot can take on ongoing responsibility, work out what needs to happen next, and come back to you for decisions that need your judgment. A dot can also review connected apps proactively and form memories from them.

OpenAI's own example captures the difference well. A dot can prepare a presentation from data and customer feedback, then keep updating the deck as new information comes in. Work finishes the deck. The dot keeps that deck up to date.

Tasks that suit a dot:

• Watching your inbox and calendar, then reminding you about missed client follow-ups.
• Keeping one project moving over several weeks.
• Turning an interview transcript into show notes and posts, then carrying your edits across all the drafts.

They can work together

Work and a dot aren't either-or. If you allow your dot to access your computer through the desktop app, the dot can create Work or Codex tasks on its own. So the dot can decide what needs doing, then hand the creation of the finished file over to Work.

Which one do you need?

• If you already know what you want made, start with Work. It's available from the Plus plan up.
• If what you need is someone who remembers, watches, and reminds, that's a dot's role. But a dot needs a Pro plan or a Business Premium seat, which cost a lot more. The plan comparison is in ChatGPT Plus vs Pro vs Business, Which One Suits a Small Team?.
• If your work is mostly tasks with a clear start and end, Plus with Work and scheduled tasks is usually enough.

One note for dots: because they keep working and sign in to work apps, take their access and approvals seriously. The guardrails to set are in Is It Safe to Sign In to Work Accounts in an AI Agent Like Dots or Grok Bot?, and who holds a dot's memory is covered in ChatGPT Now Has Dots, and What's Interesting Isn't How Smart It Is but Who Holds the Access.

Wrapping up

Work answers "please make this". A dot answers "please keep an eye on this". For most people, Work with scheduled tasks covers most needs. A dot only shows its value when there's work that needs constant watching and you're ready to set up its access properly.

If you'd like to map which work on your team suits Work, a dot, or regular automation, tell me what you need.

The photos in this article are from Unsplash, by Avtar Singh (main photo) and Kelsy Gagnebin, used under the Unsplash License.