Deep Research is available to OpenAI Plus users ($20 per month) with a limit of 10 queries per month. Pro subscribers ($200 per month) receive 120 queries and access to GPT-4.5. OpenAI positions this feature as a tool for multi-step research, which previously took hours to complete.

Key Capabilities of Deep Research

• Conducting literature reviews
• Market analysis
• Research on technologies and software
• Financial analysis
• Legal research

How Does Deep Research Work?

The Deep Research process consists of four stages:

1. Structuring the query – AI identifies key terms, sub-questions, and relevant concepts.
2. Information retrieval – Data is sourced from open databases (arXiv, PubMed, Semantic Scholar) and public resources (The Guardian, BBC, New York Times).
3. Analysis and interpretation – The system prioritizes the most important information and evaluates source quality.
4. Final report generation – A structured text is created based on the gathered data, supplemented with tables and diagrams (if needed).

Practical Use of Deep Research

To optimize the function’s performance, it is recommended to start with a query to the standard GPT-4o model, which can help refine the research request. For instance, when analyzing machine learning and energy consumption, AI may suggest additional clarifications regarding the study's context.

Throughout the process, users can monitor the addition of new sources. Upon completion (within 5 to 30 minutes), a detailed report is provided, including citations of the used sources.

Challenges and Risks of Using Deep Research

While this feature significantly speeds up information analysis, it comes with certain limitations:

• Risk of superficial analysis – The system structures data well but does not perform critical evaluation of sources.
• Reinforcement of existing biases – The algorithm may prioritize frequently cited publications, overlooking less-known but potentially significant studies.
• Quality concerns in academic work – In education, the use of Deep Research raises questions about assessing students’ independent work.

Conclusions

Deep Research by OpenAI is a powerful tool for rapid data collection and analysis, simplifying preliminary research. However, it remains in the early stages of development, with certain limitations and risks. OpenAI is expected to enhance source verification algorithms and reduce the likelihood of generating misleading conclusions.

Other companies, such as Perplexity AI and Google Gemini, are also implementing similar features, highlighting the growing role of AI in scientific research. The further development of such technologies could significantly reshape information retrieval and analysis across all fields of science and business.

Source