Join our daily and weekly newsletters for the latest updates and the exclusive content on AI coverage. Learn more
OPENAI announced today that he deploys his powerful Deep research Capacity to all Chatppt Plus,, Team,, Education And Business Users, considerably expanding access to that many experts consider the company’s most transformative IA agent from the original chatgpt.
According to an advertisement on Openai Official account xIn addition, team, education and business users will initially receive 10 depth research requests per month, while pro tier subscribers will have access to 120 queries each month.
Deep research, which is propelled by a specialized version of the next O3 modelrepresents a significant change in the way AI can help with complex research tasks. Unlike traditional chatbots that provide immediate responses, in -depth research travels independently of hundreds of online sources, analyzes text, images and PDFs and synthesizes comparable reports comparable to those produced by professional analysts.
Deep Research now takes place to all Chatgpt Plus users, Team, Edu and Enterprise?
– Openai (@openai) February 25, 2025
The AI research arms race: Deepseek’s open challenge meets the OpenAi Premium Game
The time of expanded deployment of Openai is hardly a coincidence. The generative landscape of AI has been transformed dramatically in recent weeks, with China In depth emerging as an unexpected disruptor. By open source Deepseek-R1 model under one MIT licenseThe company has fundamentally challenged the closed business model based on the subscription which defined the development of West AI.
What makes this competition particularly interesting are the divergent philosophies in play. While Optai continues to carry its most powerful capacities behind more and more complex subscription levelsDeepseek opted for a radically different approach: giving technology and leaving a thousand applications flower.
The Chinese AI Company Deepseek recently made waves when it announced R1, an open-source model of reasoning which, according to him, has achieved performance comparable to the O1 of Openai, at a fraction of the cost .
But for those who closely follow the developments of the AI, Deepseek and R1 did not come out of … pic.twitter.com/fuahyp0hz
– y combinator (@ycombinator) February 5, 2025
This strategy echoes the previous eras of the adoption of technology, where open platforms have finally created more value than closed systems. The domination of Linux in the server infrastructure offers a convincing historical parallel. For corporate decision -makers, the question becomes investing in proprietary solutions that can offer immediate competitive advantages or adopt open alternatives that could promote broader innovation in their organization.
Perplexity recent integration Deepseek -R1 in its own research tool – to a fraction of the OPENAI price – shows the speed with which this open approach can produce competing products. Meanwhile, Anthropic Claude 3.7 SONNET took another path, focusing on transparency in its reasoning process with “visible extensive reflection”.
Deepseek R1 is an impressive model, especially around what they are able to deliver for the price.
We will obviously provide much better models and it is also legitimate invigorating to have a new competitor! We are going to draw some versions.
– Sam Altman (@sama) January 28, 2025
The result is a fragmented market where each major player now offers a distinctive approach to research fueled by AI. For companies, this means a greater choice, but also increased complexity to determine the platform that is best aligned with their specific needs and values.
From the garden enclosed in the public square: Calculated democratic pivot of Openai
When Sam Altman writes this deep research “Probably worth $ 1,000 per month to certain users“, It reveals more than the elasticity of prices – it recognizes the extraordinary disparity of the value that exists among potential users. This admission is cut at the heart of the OpenAi strategic balancing act.
The company faces a fundamental tension: maintaining premium exclusivity which finances its development while simultaneously fulfilling its mission to guarantee that “general artificial intelligence benefits all humanity”. Today’s announcement represents a meticulous step towards better accessibility without undermining its income model.
I think we will first offer 10 uses per month for chatgpt plus and 2 per month at the free level, with the intention of evolving them over time.
This is probably worth $ 1000 per month for some users, but I am delighted to see what everyone in fact! https://t.co/ybicvzodpf
– Sam Altman (@sama) February 12, 2025
By limiting free level users to two monthly queries, Openai essentially offers a teaser – enough to demonstrate technology capacities without cannibalizing its premium offers. This approach follows the classic “Freemium” gaming book which has defined a large part of the digital economy, but with unusually narrow constraints which reflect the substantial IT resources required for each request for in -depth research.
The allowance of 10 monthly requests for more users ($ 20 / month) compared to 120 for professional users ($ 200 / month) creates a clear delimitation which preserves the premium value proposal. This deployment strategy on several levels suggests that OpenAi recognizes that democratizing it access to the advanced capacities of AI requires more than the simple drop in price barriers – it requires a fundamental redress in the way these capacities are packed and delivered .
Beyond the surface: the hidden forces of deep research and surprising vulnerabilities
Title figure – precision of 26.6% on “The last examination of humanity– Tells a part of the story. This reference, designed to be extraordinarily difficult, even for human experts, represents a quantum jump beyond the capacities of previous AI. For the context, the realization of 10% on this test would have been considered remarkable a year ago.
What is most important is not only raw performance, but the nature of the test itself, which requires synthesizing information through disparate areas and the application of nuanced reasoning that goes far beyond of model correspondence. Deep Research’s approach combines several technological breakthroughs: planning in several stages, recovery of adaptive information and, perhaps above all, a form of calculation self-correction which allows it to recognize and remedy its own limitations during the research process.
However, these capacities are delivered with notable dead angles. The system remains vulnerable to what could be called “consensual bias“- A tendency to favor widely accepted points of view while potentially neglecting the contrary perspectives which question established thought. This bias could be particularly problematic in the fields where innovation often emerges from conventional wisdom questioned.
In addition, the dependence of the system with regard to existing web content means that it inherits the biases and the limits of its source material. In rapid fields or niche specialties with limited online documentation, in -depth research may have trouble providing a really complete analysis. And, without access to proprietary databases or university journals based on subscriptions, its information on certain specialized areas can remain superficial despite its sophisticated reasoning capacities.
The executive dilemma: how in -depth research rewrites the rules of knowledge work
For the leaders of the C Suite C, Deep Research presents a paradox: it is a tool powerful enough to redefine the roles in all their organization, but it is still too limited to be deployed without meticulous human surveillance. Immediate productivity gains are undeniable – tasks that have required analysts’s time can now be completed in a few minutes. But this efficiency is accompanied by complex strategic implications.
Organizations that effectively integrate in -depth research will probably have to fully reinvent their information workflows. Rather than simply replacing junior analysts, technology can create new hybrid roles where human expertise focuses on framing issues, sources assessment and critical assessment of the information generated by AI. The most successful implementations will probably consider in -depth research not as a replacement of human judgment but as an amplifier of human capacities.
In -depth research for Chatgpt Plus users!
One of my favorite things that we have never shipped.
– Sam Altman (@sama) February 25, 2025
The price structure creates its own strategic considerations. At $ 200 per month for professional users with 120 queries, each request actually costs around $ 1.67 – a trivial expenditure compared to human labor costs. However, the limited volume creates an artificial rarity which obliges organizations to prioritize the questions that really deserve the capacities of deep research. This constraint can ironically lead to a more thoughtful application of technology than a purely unlimited model would encourage.
The longer term implications are deeper. While the research capacities which were once limited to elite organizations become widely accessible, a competitive advantage will derive more and more access to information, but from the way organizations supervise the questions and integrate information generated by AI on their decision -making processes. The strategic value goes from knowledge to understanding – from information to collecting information generation.
For technical leaders, the message is clear: the AI research revolution no longer comes – it’s here. The question is not to know if it is necessary to adapt, but at what speed the organizations can develop the processes, the skills and the state of mind necessary to prosper in a landscape where in -depth research has been fundamentally democratized.
