Deep Research has been the hot AI feature that everyone’s been talking about lately. If you haven’t heard about it yet, the feature lets you research “deeply” about a topic, just as the name implies. To use it, all you need to do is provide a subject for the AI to research, and then you can sit back and wait. During this time, the AI will evaluate dozens to hundreds of websites (including community forums like Reddit), extract the relevant information, and create a full-fledged, detailed report.

Gemini was the first AI tool to properly adopt the Deep Research feature, rolling it out towards the end of last year. Of course, other AI tools were quick to follow suit, with OpenAI dropping its very own Deep Research feature powered by OpenAI’s O3 model less than two months after Gemini. Similar to Gemini, OpenAI’s Deep Research is an “agentic capability” that can synthesize large amounts of information and perform multistep research tasks. Apparently, the agent can accomplish in just a few minutes what would take a human hours to do.

Like most features when initially announced, Deep Research was limited to ChatGPT Pro subscribers. OpenAI mentioned that Plus and Team subscribers would get access next. While there was initially no mention of whether free users would ever be able to access the feature, that has recently changed. Now, the tool will be available to free users as well as those subscribed to ChatGPT Plus, Team, and Pro.

Deep Research is now available for all free and paid users

As reported by TechCrunch, OpenAI shared a post on its X (formerly Twitter) account, announcing that the company is rolling out a “lightweight” version of Deep Research, powered by their O4-mini model.

OpenAI mentioned that the lightweight version is “nearly as intelligent as the Deep Research people know and love, while being significantly cheaper to serve.” As you might have guessed from the cost implications of the lightweight version, it’s targeted at free ChatGPT users and will allow the company to “increase current rate limits.” They also noted that the lightweight version’s responses will typically be shorter but will still maintain the same depth and quality as the original version.

Unfortunately, while the lightweight version is more “cost-effective” per OpenAI’s description, it doesn’t mean you’ll have free rein to use it as you please. There are still limits in place. Free users are limited to just five Deep Research queries per month. Once you hit that limit, you’ll either need to upgrade to one of the paid plans or wait until the next month to use it.

But don’t worry if you’re subscribed to the Plus or Team plan. You won’t need to settle for the lightweight version or switch to the Plus plan. OpenAI also mentioned that they’re expanding Deep Research access to Team and Pro plans. Ultimately, you’ll have access to both the original Deep Research and the lightweight, cost-effective version.

If you’re subscribed to one of the premium plans and have access to the original version of Deep Research, queries will automatically default to the lightweight version once you reach the limit, until the monthly limit resets. For paid plans, here are the limits for your reference:

• Plus & Team: 10 original Deep Research queries and 125 lightweight version Deep Research queries
• Pro: 125 original Deep Research queries and 125 lightweight version Deep Research queries
• Enterprise: 10 original Deep Research queries

What I find strange and oddly coincidental is that Gemini did pretty much the exact same thing when it launched Deep Research. It was initially only available to Gemini Advanced users, but the company rolled it out to free users just last month.

What’s even funnier is that free Gemini users can use it for five detailed reports per month, which is the same limit OpenAI has set. Coincidence or not? Guess we’ll never know!