Photo credit: www.theverge.com
OpenAI Unveils Cost-Effective Version of ChatGPT Deep Research Tool
OpenAI has launched an updated version of its ChatGPT deep research tool, claiming it to be less expensive to operate. This “lightweight” model is now accessible to both free and paid users across several subscription tiers.
The new deep research (lightweight) model utilizes a modified version of OpenAI’s o4-mini model. According to company statements, it is “nearly as intelligent” as the original deep research tool, renowned for generating comprehensive reports that include citations and a detailed overview of its analytical processes. Notably, the lightweight model is designed to produce shorter responses while still maintaining the depth and quality expected by users.
The standard deep research model is still exclusive to paid users; however, the lightweight variant is now beginning to roll out to free users, who can utilize it for up to five tasks each month. For those with paid subscriptions, the lightweight model automatically activates once users reach their limits on the standard deep research functionality. Team and Plus users will have access to a total of 25 tasks per month across both models, while Pro users can complete up to 250 tasks monthly. Meanwhile, Enterprise and Education users will be granted access the following week, with the same limitations as Team and Plus subscribers.
A critical aspect of this new model for OpenAI is its operational cost. The company has indicated that the lightweight version is “significantly cheaper” to manage. This is particularly relevant in an environment where OpenAI reportedly allocates “tens of millions of dollars” to accommodate basic user interactions, such as thanking the AI for its responses.
Source
www.theverge.com