Deep Research is not the chat box being a bit more thorough — it is a distinct mode that runs a long, multi-step investigation across many sources and returns a cited report. It takes minutes rather than seconds because it is actually reading and cross-referencing, not answering from memory. That makes it suited to questions where you would otherwise open twenty tabs: market comparisons, literature scans, due-diligence style research. The citations matter, because they let you verify the claims rather than trust them.

Scope before you run it

Because a Deep Research run is long, a vague question wastes it — it wanders across too broad a topic and returns something unfocused. Scope tightly first: state the exact question, the boundaries, and what a useful answer would let you decide. 'Research electric vans' is weak; 'Compare three electric vans under a set budget for a small delivery business, on range, payload, and total cost' is a brief it can execute well. Front-load the precision and the report comes back genuinely useful.

Frequently asked

What is ChatGPT Deep Research?

A mode that runs a long, multi-source investigation and returns a cited report, taking minutes because it actually reads and cross-references sources rather than answering from memory.

How is Deep Research different from a normal chat?

A normal chat answers quickly from the model's knowledge; Deep Research conducts a multi-step investigation across many sources and cites them, so you can verify the findings.

How do I get good results from Deep Research?

Scope the question tightly first — state the exact question, the boundaries, and the decision it should inform. A precise brief prevents it wandering across too broad a topic.