Earlier today, OpenAI announced a new ChatGPT feature, “Deep Research,” designed to help users carry out in-depth, complex information retrieval and analysis. This feature primarily targets professionals in fields such as finance, science, policy, and engineering, and also suits consumers who need detailed research, such as when buying a car, home appliances, or furniture.
▲Image source:OpenAI
OpenAI launches new “Deep Research” feature for ChatGPT
OpenAI recently announced the launch of a brand-new ChatGPT feature called “Deep Research,” designed to help users quickly complete complex web research tasks. Powered by the upcoming o3 model, this tool can perform web searches, data analysis, and multi-source integration, allowing users to obtain detailed and in-depth research reports with just a simple prompt.
▲Image source:OpenAI
The core highlight of “Deep Research” lies in its powerful multi-step reasoning and data integration capabilities. Users simply select the “Deep Research” mode in the ChatGPT input box and enter a question, and can also attach files or spreadsheets. ChatGPT then automatically searches, analyzes, and integrates multiple online sources, including web pages, images, PDF files, and more, to generate a report similar to one written by a professional research analyst. Each query may take 5 to 30 minutes, and once complete, the system notifies the user, greatly improving research efficiency.
▲Image source:OpenAI
Deep Research’s feature highlights and use cases:
- Multi-source integration: enables simultaneous aggregation of information from different websites, documents, and databases, avoiding information silos.
- Automated workflow: Automatically completes data search, filtering, and organization, saving significant manual processing time.
- Flexible applications: Suitable for market research, academic analysis, industry report writing, and even everyday consumer decisions.
- Use cases: For example, financial analysts can use Deep Research to quickly consolidate market reports from multiple investment firms; when purchasing an electric vehicle, consumers can also easily compare the performance, price, and reviews of different models.
However, Deep Research also has many challenges and limitations, such as:
- Information credibility: AI may not be able to fully distinguish authoritative data from misinformation, which could lead to information bias.
- Insufficient diversity of perspectives: The report content may lack balanced analysis of different standpoints, requiring further scrutiny by the user.
- Format limitations: Currently, only text output is supported. Data visualization and embedded images will be added in the future.
- To reduce the risk of errors, each “Deep Research” report includes detailed citations and a summary of the reasoning process, so users can verify the information. However, given the earlier controversies over the accuracy of ChatGPT’s search feature, this new feature still requires further observation and optimization.
Deep Research will first roll out to ChatGPT Pro users, with up to 100 queries per month. Support for ChatGPT Plus and Team users is expected to launch soon, with the enterprise version following shortly after. OpenAI also pledged that query limits for paid users will be “significantly increased.”
Although Google Gemini Two months agoIt also launched an AI feature of the same name, but OpenAI’s “Deep Research” in ChatGPT demonstrates stronger competitiveness in the depth of multi-source integration and citation transparency, making it especially suitable for professionals who need highly credible information.
Source: KOCPC Chinese