deep research
In plain wordsAn AI feature that spends minutes rather than seconds on a question — planning, running dozens of web searches, reading many sources — and then writes a long report with citations.
A class of agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done. products, and the task they perform: given an open-ended question, the system plans a research strategy, repeatedly issues searches, reads and filters web pages, documents or data, sometimes runs code to analyze what it finds, and synthesizes a long, structured report citing its sources, typically after 5 to 30 minutes of work. Implementations range from a single reasoning modelreasoning model, n.: An AI model trained to work through a problem step by step in hidden 'thinking' text before it answers. Letting it think longer generally makes it better at math, coding and planning. trained end to end with reinforcement learningreinforcement learning, n.: A way of teaching a computer program by trial and error: it tries actions, earns points for good results and penalties for bad ones, and gradually learns which choices pay off over time. on browsing tasks to orchestrator–workerorchestrator–worker pattern, n.: A way of organizing AI teamwork, like a newsroom: an editor AI decides which stories a big assignment needs, sends reporter AIs to cover them at the same time, and turns what they file into one report. systems in which a lead agent spawns parallel search subagentssubagent, n.: A helper AI agent that a main agent starts to handle one piece of a job; it works in its own separate workspace and reports back a short summary when it is done.. Deep research tools compress hours of desk research but inherit the weaknesses of their sources and can misattribute, overgeneralize or cite pages that do not support a claim; evaluation uses benchmarks such as BrowseCompBrowseComp, prop. n.: A test of AI web-research agents: 1,266 puzzle-like questions whose answers take a long, persistent online hunt to uncover, though each answer is short and simple to confirm. and Humanity's Last ExamHumanity's Last Exam, prop. n.: A test of 2,500 extremely hard questions with definite answers, written by about a thousand experts worldwide and billed as the last academic exam AI would need; at launch, top models scored under 10%. and expert grading of reports.
I ran deep research on battery-recycling policy and spent the afternoon checking its forty citations.
Origin
Google introduced 'Deep Research' in Gemini Advanced on 11 December 2024; OpenAI launched its own 'deep research' agent, built on a version of o3, on 2 February 2025, reporting 26.6% on Humanity's Last Exam. Perplexity, xAI, Anthropic and others released similar features during 2025, and the name became generic for the category.
Usage note
Not to be confused with research in deep learningdeep learning, n.: Machine learning done with neural networks built from many stacked layers: raw data such as pixels, sound or text goes in at one end, and each layer turns what it receives into something more abstract., or with automated AI researchautomated AI research, n.: Using AI systems to do the work of AI researchers, such as running experiments and writing code for new models, so that AI helps build its own successors., in which AI systems carry out AI research itself.
Related
- browser agentbrowser agent, n.: An AI agent that uses a web browser on your behalf — searching, clicking links, filling in forms and reading pages — to complete tasks such as booking, shopping or gathering information.
- orchestrator–worker patternorchestrator–worker pattern, n.: A way of organizing AI teamwork, like a newsroom: an editor AI decides which stories a big assignment needs, sends reporter AIs to cover them at the same time, and turns what they file into one report.
- BrowseCompBrowseComp, prop. n.: A test of AI web-research agents: 1,266 puzzle-like questions whose answers take a long, persistent online hunt to uncover, though each answer is short and simple to confirm.
- WebGPTWebGPT, prop. n.: A 2021 OpenAI research system that answered questions by searching and browsing the web in a simple text browser, collecting sources and citing them — an early ancestor of today's research agents.
- citationcitation, n.: A pointer from something an AI says to the source that supports it — a link, document or quoted passage — so people can check the claim for themselves.
- reasoning modelreasoning model, n.: An AI model trained to work through a problem step by step in hidden 'thinking' text before it answers. Letting it think longer generally makes it better at math, coding and planning.
- agentic RAGagentic RAG, n.: Retrieval in which the AI runs its own research: it decides what to search for, looks at the results, searches again if needed, and only then answers.
- retrieval-augmented generationretrieval-augmented generation, n.: Giving an AI an open-book exam: before answering, it looks up relevant passages in a set of documents and bases its reply on them, often citing where each point came from.
Sources
- Google (2024). Try Deep Research and our new experimental model in Gemini, your AI assistant. The Keyword, 11 December 2024. blog.google/products/gemini/google-gemini-deep-research
- InfoQ (2025). OpenAI Launches Deep Research. 6 February 2025. infoq.com/news/2025/02/deep-research-openai
- Anthropic (2025). How we built our multi-agent research system. anthropic.com/engineering/multi-agent-research-system
Mentioned in
AI agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done., agentic RAGagentic RAG, n.: Retrieval in which the AI runs its own research: it decides what to search for, looks at the results, searches again if needed, and only then answers., AI scientistAI scientist, n.: An AI system meant to do research largely on its own: reading the literature, coming up with ideas, running experiments and writing up the results., BrowseCompBrowseComp, prop. n.: A test of AI web-research agents: 1,266 puzzle-like questions whose answers take a long, persistent online hunt to uncover, though each answer is short and simple to confirm., browser agentbrowser agent, n.: An AI agent that uses a web browser on your behalf — searching, clicking links, filling in forms and reading pages — to complete tasks such as booking, shopping or gathering information., citationcitation, n.: A pointer from something an AI says to the source that supports it — a link, document or quoted passage — so people can check the claim for themselves., GAIAGAIA, prop. n.: A test of AI assistants made of real-world questions that people can answer with a browser and some patience, but that require an AI to search, read files and combine facts., orchestrator–worker patternorchestrator–worker pattern, n.: A way of organizing AI teamwork, like a newsroom: an editor AI decides which stories a big assignment needs, sends reporter AIs to cover them at the same time, and turns what they file into one report., question answeringquestion answering, n.: Systems that reply to people's questions with the right answer, whether by finding it in a passage they are given, searching a database or the web, or drawing on what they already know., agentic reinforcement learningagentic reinforcement learning, n.: Training an AI assistant by rewarding success on long, multi-step jobs where it must use tools, browse, run code and react to what happens, rather than answer in one go., AI co-scientistAI co-scientist, n.: An AI system built to work alongside human scientists, proposing and debating new research ideas and hypotheses that people then test in the lab., WebGPTWebGPT, prop. n.: A 2021 OpenAI research system that answered questions by searching and browsing the web in a simple text browser, collecting sources and citing them — an early ancestor of today's research agents.