OpenAI appears to be on the verge of expanding its toolset for ChatGPT users with the upcoming release of a new model variant dubbed “Deep Research Mini.” This lightweight version of the existing Deep Research capability is now surfacing within the ChatGPT interface under a model selector labeled "?model=research", signalling that it may already be accessible to some users as part of a gradual rollout.
Update: Rolling out to all users 👀
The lightweight version of deep research is powered by a version of OpenAI o4-mini and is nearly as intelligent as the deep research people already know and love, while being significantly cheaper to serve.
— OpenAI (@OpenAI) April 24, 2025
Responses will typically be shorter while maintaining the depth and… pic.twitter.com/H2UD5GThVj
The feature is reportedly designed to bring some of the analytical depth of the premium Deep Research model to a broader audience, including those on the Free tier. While full details remain sparse, early descriptions suggest it will allow users to obtain in-depth, topic-specific breakdowns, ranging from consumer decision-making to educational guidance, without the need for a premium subscription.
the lightweight version of Deep Research in ChatGPT seems to be getting prepared for release
— ʟᴇɢɪᴛ (@legit_api) April 24, 2025
Free users and above will get access to this! pic.twitter.com/FB92RqAtz7
"Free users now have access to the lightweight version of deep research, so you can get detailed insights on any topic, from shopping to choosing the right college."
Try Deep Research Mini (Consumes Deep Research Quota) 👇
The name “Deep Research Mini” reflects its likely positioning as a trimmed-down version of the more computationally intensive Deep Research model, which has previously been reserved for Pro users. This aligns with OpenAI’s broader trend of introducing tiered capabilities, offering users a taste of advanced functionality while preserving the full experience for paid tiers.
OpenAI, known for its GPT model family and the consumer-facing ChatGPT product, has frequently iterated on how it delivers advanced tools to users. Making research-level capabilities accessible to Free users would mark a notable shift, hinting at broader ambitions to democratize access to sophisticated AI reasoning without placing it behind a paywall. As of now, it’s unclear when the lightweight Deep Research mode will be fully rolled out or whether it will retain all the contextual depth of its full counterpart.
