Following the recent release of the o3 mini model, OpenAI has introduced an innovative ChatGPT agent known as ‘deep research.’ This new addition utilizes an upcoming version of OpenAI’s reasoning model, o3, enabling it to execute multi-step research on the internet for complex tasks. The deep research AI in ChatGPT essentially serves as a personal research assistant, tailored to support users engaged in intensive knowledge domains such as finance, science, policy, and engineering. However, its versatile capabilities also make it suitable for any detailed research demands, such as investigating options for a new car purchase.
Table of Contents
Competing Innovations in AI Research
The concept of deep research AI is reminiscent of Google’s Gemini, which launched a similar tool under the same name in early December. The ongoing competition between these tech giants emphasizes the advancement of AI technologies, particularly models designed to enhance reasoning capabilities and execute tasks on behalf of users. Simultaneously, the emergence of DeepSeek, a China-based AI, with its cost-effective R1 model, has jolted the tech sector, further intensifying the rivalry in artificial intelligence.
OpenAI’s Strategic Direction
OpenAI’s CEO, Sam Altman, acknowledged the impressive capabilities of DeepSeek’s R1 model, particularly regarding its affordability. He expressed enthusiasm about the competitive landscape, stating, ‘We will obviously deliver much better models, and it’s invigorating to have a new competitor!’
Continuing on their path, OpenAI plans to enhance their research roadmap, emphasizing the ever-increasing need for computational power to achieve their mission of developing artificial general intelligence (AGI). Altman’s announcements suggest a commitment to releasing products to keep users engaged and satisfied.
Deep Research’s Unique Features
When comparing deep research AI to other models like GPT-4o, which excels in real-time conversational tasks, the deep research tool stands out for its focus on thorough and specific projects, requiring timeframes that can range from five to thirty minutes. Users can confidently step away from their devices while deep research handles the task, receiving a notification when the research report is complete.
The outputs are comprehensive, providing detailed summaries and insights into the research process, complete with citations and links to original sources.
Availability and Future Integration
At launch, the deep research feature is exclusively accessible for ChatGPT Pro users with a limit of 100 queries per month. Following this, the availability will extend to Plus and Team users, and eventually to Enterprise users. Currently, deep research is only on the web, with mobile and desktop application access expected by the end of February. Users in the UK and EU may need to wait a bit longer, as OpenAI continues to work on providing access to these regions.
Frequently Asked Questions
What is OpenAI’s deep research AI?
OpenAI’s deep research AI is a new feature in ChatGPT that conducts multi-step research on the internet for complex tasks, acting as a personal research assistant.
Who can use the deep research AI?
Initially, deep research AI is available to ChatGPT Pro users, with plans to roll it out to Plus, Team, and Enterprise users thereafter.
How does deep research differ from other ChatGPT models?
Unlike models like GPT-4o that focus on real-time conversations, deep research is optimized for detailed, specific tasks, taking anywhere from five to 30 minutes to complete.
What features does deep research AI offer?
Deep research AI offers comprehensive outputs and summaries, including citations and links to sources, and allows users to step away while it works.
When will deep research be available on mobile and desktop apps?
Deep research is set to be integrated into mobile and desktop applications by the end of February.
