Chat, Work, and model choices

Home Tools Platforms ChatGPT Chat, Work, and model choices

Choose an interaction style and compare model or reasoning settings against a clear task.

Conversation & research · Documented · Documentation checked 2026-10-04

ChatGPT Chat and Work; local Codex and the OpenAI API are separate surfaces.

Tools, usage, and controls vary by plan, client, region, rollout, and workspace policy.

The worked example is illustrative and has not been tested in a signed-in product account.

What it does and how it connects

Choose an interaction style and compare model or reasoning settings against a clear task.

Chat is useful for an explanation or a quick exchange. Work starts from an outcome and can use context and tools to make a deliverable. A model is the engine producing the response; reasoning effort is a processing choice, not a permission to access more data. Available choices depend on your client and account.

Walkthrough

Use the product named in the access panel. The steps below combine documented controls with the editorial example.

  1. Start a new chat with a small question you can independently answer.
  2. For a deliverable, choose Work if available and state the output format and success criteria.
  3. Inspect the available model or reasoning control. Keep the same input when comparing two settings.
  4. Review the answer and ask one targeted follow-up; save the version you actually checked.

Worked example

Imagine planning a community AI workshop. Ask for a beginner explanation in Chat, then use Work to turn your approved outline into a session plan. The teaching goal stays constant while the expected output changes. This is an illustrative exercise, not a measured model comparison.

Explain the difference between a model, an assistant, a workflow, and an agent to a curious beginner. Use one community-workshop example for all four. Then propose a 30-minute lesson plan. Label assumptions and tell me which parts need current sources. Do not connect accounts or publish anything.

Inspect and verify the result

Check that the four terms remain distinct. A workflow follows an arranged process; an agent may select tools and intermediate steps. Judge the lesson against your time limit and audience, not how confident the answer sounds. Compare factual accuracy before comparing style or speed.

Limits and troubleshooting

If a control is missing, check the documented plan and surface rather than assuming an outage. More reasoning can still produce a wrong answer. Usage is shared or metered differently across products; a ChatGPT subscription does not establish identical API access or pricing.

Data, permissions, and actions

The prompt and supplied context enter the conversation. Work may invoke enabled tools, so name permitted sources and actions. Choosing a model changes neither connected-account permissions nor organizational access policy.

Search and deep research

Find current information and build a report whose sources can be checked.

Explore Search and deep research

From evidence to a beginner lesson

Create a sourced workshop lesson and an editable handout from approved material.

Explore From evidence to a beginner lesson

A weekly evidence digest with review

Repeat a public-source investigation and deliver a private draft digest.

Explore A weekly evidence digest with review

Evaluations: does the system do the job?

Create representative tests for quality, tool behavior, refusal, escalation, and the effects of a change.

Explore Evaluations: does the system do the job?

Sources and verification

Documentation checked 2026-10-04. Examples are illustrative unless described otherwise.

Using ChatGPT: Chat and Work · OpenAI

Model selection · OpenAI

Usage and billing · OpenAI

Share this page