the30top
CH

ChatGPT Deep Research

by OpenAI

Multi-step web research that writes a cited report

Reviewed by the the30top research team · Last reviewed 2026-10-27

Overview

OpenAI's deep research does multi-step research on the internet and produces a documented report with citations to specific passages in its sources, and it can also work over files you upload. Runs take roughly 5 to 30 minutes. OpenAI states that it can hallucinate facts, struggle to separate authoritative sources from rumor, and misjudge its own confidence, so every report needs human checking.

Our verdict

A capable general research agent with a free entry point and clear plan tiers. Treat output as a strong first draft that needs verification, not a finished deliverable.

Best for

Analysts and consultants who want a general purpose research agent on a per-user subscription.

Avoid if

You need research grounded in licensed financial or expert-network content.

Pros

  • Reports cite specific passages from sources
  • Free plan includes limited access
  • Handles uploaded files alongside web sources

Cons

  • Runs can take up to 30 minutes
  • Can hallucinate and misjudge confidence
  • Does not reach licensed financial databases

Key capabilities

Multi-step research across hundreds of web sourcesCitations to specific passages in sourcesWorks over user uploaded filesCharts built with the python toolSummary of the reasoning behind each report

Frequently asked questions

Which ChatGPT plans include deep research?

The pricing page shows limited access on Free and Go, expanded on Plus at $20 per month, and maximum on Pro from $100 per month.

How long does a deep research run take?

OpenAI says it may take anywhere from 5 to 30 minutes.

Can deep research make mistakes?

Yes. OpenAI says it can hallucinate facts, make incorrect inferences and fail to convey uncertainty well.

More AI Research Agents (deep research)