OPENAI expands the deep exploration to plus users by heating AI agents wars with Deepseek and Claude

OPENAI announced today that it unfolds its powerful Deep research ability to all Chatgpt plus., Team., Education and Entity Consumers who significantly expand access to what many experts consider the most transformative agent of the company of the company after the original Chatgpt.

According to a message about OpenAi’s Official X AccountPlus, team users, education and businesses will initially receive 10 deep research requests per month, while Pro -tiier Acrowners will have access to 120 requests a month.

Deep Research, which is powered by a specialized version of the upcoming OPENAI O3 modelIt is a significant change in how AI can help with complex research tasks. Unlike traditional chatbots that provide immediate answers, deep research independently searching hundreds of online sources, analyzing text, images and PDFs and synthesizes comprehensive reports comparable to those made by professional analyzers.

Deep Research is already deploying all users of Chatgpt Plus, Team, Edu and Enterprise?

– Openai (@openai) February 25, 2025

The Ai Research Arms Race: Deepseek Open Challenge meets the first -class Openai game

The time of the extended performance of Openai is hardly accidental. AI’s generative landscape has been drastically transformed in recent weeks with China Deepseek arising as an unexpected destroyer. By an open source of them Model Deepseek-R1 underne My licenseThe company has disputed a radically closed business model based on a subscription that determined the development of Western AI.

What makes this competition particularly interesting is the different philosophies in the game. While Openai continues to complete its most powerful capabilities behind ever -increasing Subscription levelsDeepseek has chosen a radically different approach: give out the technology and let a thousand applications bloom.

The Chinese AI company Deepseek recently made waves when it announced R1, an open source reflection model that claimed to have achieved a comparable performance with O1 on Openai, with part of the price.

But for those who follow a close AI, Deepseek and R1, they didn’t get out of … pic.twitter.com/fuahyp0hz

– Y Combine (@ycombinator) February 5, 2025

This strategy sounds the larger eras for adopting technology, where open platforms ultimately create more value than closed systems. The dominance of Linux in the server infrastructure offers an insurmountable historical parallel. For persons making Enterprise decisions, the question is whether to invest in their own decisions that can offer immediate competitive advantages or cover open alternatives that could encourage broader innovations in their organization.

Perplexity Recent integration On Deepseek-R1 in its own research instrument, part of the Openai-Deomontized price point is how quickly this open approach can give competitive products. Meanwhile, anthropic Claude 3.7 Sonnet has taken over one more time, focusing on transparency in the process of his reasoning with “visibly expanded thinking.”

Deepseek R1 is an impressive model, especially around what they can deliver for the price.

Obviously we will deliver much better models, and it is also legitimate to have a new competitor! We will take out several editions.

– Sam Altman (@sama) January 28, 2025

The result is a fragmented market where each major player now offers a distinctive approach to AI -powered research. For enterprises, this means a greater choice, but also increased complexity in determining which platform is best arranged in accordance with their specific needs and values.

From garden with walls to public square: the calculated democratic rotation of Openai

When Sam Altman writes this deep study. “Probably cost $ 1,000 a month to some users“He reveals more than simple price elasticity – he recognizes the exceptional value of the value that exists among potential users. This recognition reduces the heart of the continuing act of strategic balancing OPENAI.

The company is facing a basic tension: maintaining the premium exclusivity that finances its development while fulfilling its mission to ensure that “artificial general intelligence is beneficial for all humanity.” Today’s message is a careful step to more accessibility without undermining its revenue model.

I think we will initially offer 10 uses per month for Chatgpt Plus and 2 per month in the free layer, with the intention of scale them over time.

It probably costs $ 1,000 a month to some users, but I’m excited to see what everyone is doing with it! https://t.co/ybicvzodpf

– Sam Altman (@sama) February 12, 2025

By restricting users to a free level to just two requests a month, Openai essentially offers a teaser – enough to demonstrate the capabilities of the technology without canniblyizing their premium proposals. This approach follows the classic book “Freemium”, which defines a large part of the digital economy, but with unusually strict restrictions that reflect the essential computing resources needed for each deep research request.

Distribution of 10 monthly requests for plus users ($ 20/month) compared to 120 for Pro users ($ 200/month) creates a clear outline that retains the premium value proposal. This multi -stage launch strategy suggests that Openai acknowledges that democratization of access to advanced AI capabilities requires more than simply a decrease in price barriers – it requires a fundamental rethinking of how these capabilities are packaged and delivered.

Beyond the surface: the hidden strengths of deep research and surprising vulnerabilities

The title figure – 26.6% accuracy of “Humanity’s last exam” – says only part of the story. This indicator, designed to be extremely challenging even for human experts, is a quantum jump beyond previous AI options. For context, achieving even 10% of this test would be considered remarkable only a year ago.

The most important is not only the raw performance, but the nature of the test itself, which requires the synthesis of information in different domains and the application of nuanced reasoning that exceeds the coincidence of the model. The Deep Research approach combines several technological breakthroughs: multi-stage planning, adaptive information retrieval and, most importantly, a form of computing self-decoration that allows it to recognize and eliminate its own restrictions during the examination process.

Still, these options come with remarkable blind spots. The system remains vulnerable to what can be called. ”Consensus” – a tendency for privileges widely accepted perspectives, while neglecting the contrasting perspectives that dispute established thinking. This addiction can be especially problematic in areas where innovation often emerges from challenging conventional wisdom.

Moreover, reading the system from existing web content means that it inherits the biases and restrictions of its output material. In rapidly developing fields or niche specialties with limited online documentation, deep research can fight to provide truly comprehensive analysis. And without access to their own databases or academic magazines -based subscription, its information on certain specialized domains can remain superficial, despite its complex possibilities for reasoning.

Executive Power Dilemma: How Deep Studies Rewrite Knowledge Rules

For the leaders of the C-Suite Deep Research, it is a paradox: it is a tool powerful enough to redefine roles in their entire organization, but is still too limited to be implemented without careful human supervision. Immediate productivity profits are indisputable – tasks that once require days of analyzer, can now be completed in minutes. But this effectiveness comes with complex strategic consequences.

Organizations that effectively integrate deep research will probably have to rethink their information work processes. Instead of simply replacing junior analysts, technology can create new hybrid roles in which human expertise focuses on framing issues, evaluating sources and critical evaluation of AI insights generated. The most successful realizations are likely to consider deep research not as a substitute for human judgment, but as an amplifier of human capabilities.

Deep study for Chatgpt Plus users!

One of my favorite things we have ever delivered.

– Sam Altman (@sama) February 25, 2025

The price structure creates its own strategic considerations. With $ 200 a month for Pro users with 120 requests, each request effectively costs about $ 1.67 – trivial costs compared to human labor costs. However, limited volume creates an artificial shortage that forces organizations to prioritize which issues really deserve the opportunities of Deep Research. This restriction can ironically lead to a more thoughtful application of technology than a purely unlimited model would encourage.

The longer-term consequences are deeper. As the research capabilities that have once been limited to elite organizations become widely available, the competitive advantage will increasingly stem not from access to information, but from how organizations create questions and integrate AI -generated insights for their decision -making processes. Strategic value is shifted from knowing to understand – from gathering information to the generation of insight.

For technical leaders, the message is clear: the AI ​​Research revolution is no longer coming – it’s not. The question is not to adapt, but how quickly organizations can develop the processes, skills and cultural way of thinking needed to flourish in landscape where deep research was democratized fundamentally.