OpenAI has introduced a lightweight version of its deep research tool for ChatGPT, named o4-mini, available to all users including those on the free tier. This version activates once users reach their usage limits and provides shorter but thorough answers, with an accuracy of 45.6% compared to the full deep research's 51.5%. The o4-mini model was made available starting April 24. The new o4-mini model is significantly less costly to serve than the full deep research tool, allowing OpenAI to extend its usage to more users. For Plus, Team, and Pro users, the rate limit is set at 25 and 250 monthly tasks respectively, while free users can access up to five answers from o4-mini. Enterprise and Education tier users will have access to o4-mini next week, following the same usage rules as Team users. This development follows a trend among major AI companies to integrate reasoning models into their chatbots, with similar offerings from Anthropic, Google, and xAI. OpenAI announced the introduction of o4-mini on X, stating it is part of their strategy to balance increased usage with more cost-effective offerings.