After OpenaiThe confusion of the AI joins the “Deep Research” Beagwire. And it does it in a rather interesting way. In the footsteps of episodes Deepseek‘S “Argumenting” model, confusion is the first large AI provider that also offers a deep research function for free users. In comparison Openais Deep Research Feature is only available in the subscription of $ 200 per month.
Deep Research is an upcoming AI function that takes a little time, but carries out dozens of relatives search processes, runs over hundreds of resources and uses an argumentation model to logical every input request in a step-for-step process. You can do similar results from tools like like Copilots “think deeper” Safe, but what distinguishes deep research is that all information you have summarized are summarized in a comprehensive report on white paper style.
Deep Research is free of charge for all registered users, although they are limited to just five queries per day. However, if you pay for confusion professional (which costs 20 USD per month or $ 200 per year), you will receive up to 500 questions per month. Openais Deep Research feature is initially limited to 100 inquiries per month.
Although it is not a apples-to-color comparison. The deep research function of Chatgpt takes a much currently. Up to 20 minutes. Requested questions are asked, all the complex steps that are connected in his process are displayed and ultimately spits out a very long report. Confused deep research feature is a kind a Lite Version of it. You will receive an answer in 2-4 minutes, so your results are of course dependent on the search for the AI for web and data consumption as a deep interpretation. Openai uses its upcoming O3 argumentation model for deep research, but confusion has not mentioned any details about the model it uses.
Credit: confusion
In The last exam of humanityA frequently used AI benchmark, which consisted of over 3,000 questions in a number of topics, achieved Deep research of the confusion of 21.1%for the accuracy, which is much higher than Deepseek R1 (8.6%) and Gemini (7.2%). Openai’s deep research still has the lead with a degree of 26.6%, but here a silver medal is respectable given the much lower entry barrier of the tool.
How does this change how you could use confusion? So far, AI chatbots have only participated in several input requests. You ask a question and keep asking them to get detailed answers. But with deep research you can ask a single question and be ready. The more specific and more detailed your prompt, the better the research and the report of the bot will be, but the AI can now give you pages with information in response to much less request. And as soon as your report has been created, you can download it as a PDF.
Credit: confusion
Over in his blogConfusion has emphasized several examples of the type of differences that can make deep research. If the standard model may have an overview of bullet point answers, deep research will instead return several paragraphs, detailed reports and extended formatting.
