Skip to main content
Research

OpenAI Deep Research

Conduct comprehensive research using AI agents that autonomously search the web, analyze data, and synthesize findings into detailed reports with citations.

View details
Try it in Ampere

Inputs

Loading input fields...
Execution Steps

Loading workflow structure...

Loading curated examples...

Overview

OpenAI Deep Research turns a detailed research question into a longer cited report. Use it for market scans, competitor context, category explainers, evidence gathering, and strategy briefs when you need multi-step web research, synthesis, and source review rather than one quick answer.

Use cases

  • Research a market, category, or competitor set before writing a positioning or campaign brief.
  • Gather cited context for a product launch, founder narrative, or buyer-education article.
  • Ask for a structured report with sections, comparisons, and source-backed findings.
  • Use code-enabled research when the task may need calculations, data analysis, or tables.

Input tips

  • Write a specific query with the market, audience, region, timeframe, and output structure you need.
  • Use systemInstructions for focus areas, citation expectations, or report-format preferences.
  • Choose o4-mini-deep-research for the default lighter path or o3-deep-research for deeper synthesis.
  • Leave prompt enrichment on for brief or vague queries; turn it off for already-detailed prompts.
  • Use maxToolCalls from 50-500 to bound research depth and runtime; 100 is the default.
  • Disable the code interpreter when the task is text-only research.

Expected output

The AI Tool returns a synthesized Markdown research report, validated citations with source titles and URLs, citation-validation stats, model and usage metadata, optional research-activity timeline, optional reasoning summary, and cost metadata. The output view supports rendered/raw report views, copy, Markdown download, citation browsing, and full-screen reading.

Caveats

  • Deep research can take several minutes and may take much longer for broad or complex questions.
  • Sources, search results, and public web availability can change between runs.
  • Citations and findings should be reviewed before using them in public claims or strategy documents.
  • A lower maxToolCalls value can reduce runtime but may limit research depth.
  • The AI Tool synthesizes research; it does not guarantee exhaustive coverage or definitive market truth.

Related AI Tools