Watch the team that built it run real queries so you see what a good brief looks like.
Introduction to Deep Research (launch demo from Tokyo)
OpenAI Feb 2025
Watch on YouTube youtube.com →Deep research modes in ChatGPT, Gemini, Perplexity and Claude are agents: you give a detailed brief, approve the research plan, and 5 to 30 minutes later get a long cited report built from dozens or hundreds of sources. The quality depends almost entirely on your brief, so specify scope, geography, timeframe, output structure and what sources to prefer. Treat the report as a strong first draft from a junior analyst and spot-check the claims that matter.
19 resources.
Watch the team that built it run real queries so you see what a good brief looks like.
OpenAI Feb 2025
Watch on YouTube youtube.com →The fastest way to grasp the plan-browse-synthesize loop before you spend a query.
YouTube explainer Feb 2025
Watch on YouTube youtube.com →The researcher who built deep research explains how it thinks, which makes you a better prompter.
No Priors (Sarah Guo, Elad Gil) Jun 2025
Watch on YouTube youtube.com →A same-prompt shootout so you can judge output quality differences yourself.
YouTube AI reviewer May 2025
Watch on YouTube youtube.com →Isa Fulford and Josh Tobin on what the agent is actually good and bad at, from the source.
Sequoia Capital Apr 2025
Open sequoiacap.com →The primary source that defined the deep research category, with benchmark data and honest limitations.
OpenAI Feb 2025
Open openai.com →The official how-to including query limits per plan, so you spend your monthly runs wisely.
OpenAI Mar 2025
Open help.openai.com →The free-tier option: near report quality in under 3 minutes, ideal for bootstrapped founders.
Perplexity Feb 2025
Open perplexity.ai →Google's launch post explaining the plan-approve-refine flow Gemini pioneered.
Google Dec 2024
Open blog.google →Short official playbook on editing the research plan and asking for report revisions.
Google Mar 2025
Open blog.google →One page on capabilities and supported inputs, useful before committing to a tool.
Google Apr 2025
Open gemini.google →A widely shared operator thread: use a strong model to write your deep research prompt first.
BuccoCapital Bloke Feb 2025
Open x.com →A fill-in-the-blanks system that reliably turns vague questions into report-grade briefs.
Ahmad Osman Feb 2025
Open ahmadosman.com →A startup operator runs the same $5M ARR brief through both tools and shows the raw outputs.
Greg Isenberg Feb 2025
Open gregisenberg.com →A prompting-focused comparison that shows how brief quality changes the winner.
Learn Prompting Feb 2025
Open newsletter.learnprompting.org →A clear practitioner explainer of the five-step pipeline behind the reports.
Gend Apr 2025
Open gend.co →A blunt business-lens review that saves you from trusting the wrong tool for client-facing work.
Section (AI business school) Feb 2025
Open sectionai.com →For technical founders: wire deep research into your own product or internal tools.
OpenAI Jun 2025
Open developers.openai.com →A living library of proven deep research prompts across OpenAI, Gemini and Qwen.
LangGPT community Mar 2025
Open github.com →The same ground, over in Understand your customers, our Starting Up track.