Introducing deep research

Source: OpenAI Date: 2025-02-02 URL: https://openai.com/index/introducing-deep-research

Summary

Summary

OpenAI launched Deep Research in February 2025 — a ChatGPT feature that conducts extended autonomous research tasks, spending minutes to hours browsing the web, reading documents, synthesizing information, and producing detailed research reports. Deep Research uses an o3-based reasoning model to plan and execute multi-step research workflows rather than producing a single response.

Implications

Agentic engineering/product thread. Deep Research is the research workflow equivalent of Operator for web browsing — it’s an autonomous agent product deployed for a specific high-value use case. The multi-hour research framing signals a qualitative shift in what AI can do: not answer a question, but complete a research project. The February 2025 timing places Deep Research in the opening salvo of the agentic product era, alongside Operator. For knowledge workers — analysts, researchers, consultants — Deep Research directly addresses the most time-consuming part of their work. The competitive response from Perplexity and Google’s similar features demonstrates how quickly the capability category was validated and replicated.