All free users of the AI chatbot may now access ChatGPT’s Deep Research mode, which was previously only available to subscribers who paid, allowing them to improve their research skills with more thorough answers.
Lightweight Mode For Deep Research
The move by OpenAI to make Deep Research available to free users marks a substantial increase in ChatGPT’s free tier functionality. Five research projects per month are available to non-paying customers with this lightweight edition, which is powered by the economical o4-mini model. The capability was sent out to all free users by OpenAI over a period of days starting in late April 2025.
This growth is in line with OpenAI’s overarching plan to democratize cutting-edge AI capabilities while preserving premium incentives for premium subscriptions. Previously, other paid features like Advanced Voice Mode and image production (only a few photographs each day) were only partially accessible to free users. OpenAI maintains significant advantages for Plus, Team, and Pro subscribers, who receive significantly higher usage limits and access to the more potent o3-powered version, while balancing competitive pressure from rivals by providing its free user base with a taste of Deep Research’s potent multi-step reasoning and web browsing capabilities.
Comparing Paid Features
Deep Research is now available to free users, although there are some significant restrictions in comparison to paying tiers. While Plus and Team customers receive 10 tasks per month using the full o3-powered version plus an extra 15 tasks with the lightweight version, the lightweight edition only delivers five tasks per month. The most generous allowance is given to pro subscribers, who complete 125 assignments per month for each version.
There are notable variations in capabilities beyond consumption constraints. The o4-mini model, which powers the free lightweight version, produces responses that are shorter than those of the more expensive o3 model while retaining a similar level of quality. Additionally, paid customers have access to deeper reasoning models, greater communications and file upload limitations, priority access to OpenAI’s newest features, and restricted access to Sora video production. The $200 Pro subscription offers significantly more research capability for professionals who depend on it, which makes it beneficial for people in knowledge-intensive areas like finance, science, and law.
Capabilities Of The Deep Research Mode
The most sophisticated agentic capacity offered by OpenAI is Deep Research, which does multi-step research on its own by searching the web, evaluating data, and compiling findings into in-depth reports. It can do tasks that would take people several hours in tens of minutes thanks to modified versions of OpenAI’s models (o3 for the full version and o4-mini for the lightweight version). The program is especially useful for knowledge workers in the fields of law, science, and finance because it is excellent at locating specialized, counterintuitive information across a large number of websites.
Users can include spreadsheets, files, or photos with their Deep Research queries to give context. Before starting its study, the system could create forms to record particular parameters. Research sessions normally take 5-30 minutes to complete, and a sidebar shows the model’s thought process and the websites it consults. A crucial characteristic that sets Deep Research apart from typical ChatGPT discussions is the clear citations to sources included in all outputs, which make information verification simple. This knowledge synthesis capability is a major step toward OpenAI’s overarching objective of creating artificial intelligence (AGI) that can generate original scientific research.
Simplified Research Instrument
Built on OpenAI’s o4-mini paradigm, the lightweight version of Deep Research maintains essential functionality while maximizing efficiency and accessibility. Compared to the full o3-powered version, this condensed version generates shorter responses while retaining key features like multi-step reasoning, real-time web browsing, and document parsing. Notwithstanding these variations, OpenAI asserts that the lightweight model provides comparable levels of analytical quality and depth, making it appropriate for challenging research projects and allowing the business to increase access for all user tiers.
In response to the strong demand for Deep Research capabilities, OpenAI launched this lightweight substitute, which enables the firm to expand access for free users while raising usage caps for paid members. Users of Plus, Team, and Pro automatically switch to the lightweight version for more questions once they have completed their monthly allotment of complete Deep Research projects. While still offering its full user base useful research capabilities, this tiered approach aids OpenAI in managing the computational expenses related to these resource-intensive procedures.