OpenAI has unveiled an updated version of its ChatGPT deep research tool, which is designed to be more cost-effective. The new “lightweight” model is now accessible to both free and various paid users of the platform.
This lightweight variant is based on OpenAI’s o4-mini model and is said to be “nearly as intelligent” as the original deep research tool, which is capable of generating extensive reports along with citations and a thorough explanation of its findings. According to the company, the lightweight model provides responses that are more concise but still uphold the standard of depth and quality users expect.
While the original deep research tool remains available only for paid subscribers, the lightweight version is being rolled out to free subscribers, who can utilize it for up to five tasks each month. Paid users will also be able to access the lightweight model, which will be automatically activated when they reach their usage limits on the original version. Team and Plus users will have a total of 25 monthly tasks across both models, while Pro users will have 250. Enterprise and Education users are expected to gain access next week, with the same limits as Team and Plus users.
OpenAI has emphasized that this new version is “significantly cheaper” to operate, a notable advantage given the company’s expenditure of “tens of millions of dollars” on processing user interactions such as saying please and thank you.
