Unlock OpenAI's Free 'Think Deeper' Feature in Microsoft Copilot

Windows enthusiasts and productivity buffs, you’re in for a treat: OpenAI's high-performing "o1 Reasoning Model," lovingly nicknamed the "Think Deeper" feature, is now free for all Microsoft Copilot users. If you’ve been following the AI wars (yes, they exist, folks), you may already suspect that this news reeks of a strategic counter-move aimed at the recent buzz around Deepseek, a Chinese AI competitor. But we’re not here solely for drama—there’s a lot more under the hood to unpack. Let’s break it down and see why this matters to you (whether you’re casually talking to an AI chatbot or diving into complex programming).

---

What Just Happened?​

OpenAI, in collaboration with Microsoft, just ripped the subscription gate off their o1 model for Copilot users. Previously, you’d have to pay between $20 to $200 per month to access this advanced reasoning capability via OpenAI’s platform. This sudden shift comes days after Deepseek, another AI company, rolled out its own reasoning model called R1.
The kicker? Microsoft's CEO of AI, Mustafa Suleyman, announced that Copilot’s "Think Deeper" button is now available without requiring a special Copilot+ or ChatGPT subscription. Simply put, whether you’re using Copilot inside Word or Excel (or integrating it elsewhere across Microsoft 365), you now have a new AI superpower at your fingertips—all for free.
No Pro edition, no hidden paywalls. This is groundbreaking as it democratizes some of the best reasoning tech available in natural language processing today.

---

How “Think Deeper” Works (And Why You Should Care)​

At its core, the “Think Deeper” button in Copilot extends the time the AI spends processing your question. Typically, many AI models, including GPT-4, prioritize efficiency—delivering a response quickly means sacrificing the depth of reasoning. OpenAI’s o1 model flips this convention.
Here’s why the o1 model stands out:

• Extended Processing Time: Think of it as telling your AI to take a few deep breaths and mull it over for 30 seconds. This allows it to refine its understanding and structure more thoughtful responses.
• Better STEM Problem Solving: OpenAI claims their o1 reasoning model solves 83% of International Mathematics Olympiad problems. In comparison, a standard GPT-4 adaptive response tops out at solving only 13%. That chasm is massive. Whether you’re crafting a Python algorithm or debugging trigonometry equations, the o1 model is a game-changer.
• Explicit Reasoning Steps: Unlike its predecessors, o1 walks you through the steps it took to arrive at an answer. You’re not just getting answers; you’re seeing the thought process—critical for STEM fields and complex problem-solving.
• Offline Reasoning: Unlike some competitors, the o1 feature doesn’t access the internet during "Think Deeper" sessions. While this means zero real-time data feeds, it also enhances focus by ensuring reasoning remains self-contained (and arguably more secure).

This makes the o1 feature ideal for anyone needing detailed, accurate, and step-by-step reasoning—be it students, tech professionals, or anyone diving deep into logic-heavy tasks.

---

Up Close: Why This Move Smells of AI Chess Mastery​

Deepseek’s Rise and Microsoft’s Counter​

Deepseek's R1 model had a blockbuster week, receiving accolades for supposedly matching o1’s reasoning capabilities. OpenAI’s decision to waive its pricing imposition isn’t just consumer-friendly—it’s surgical. Deepseek is already mired in controversy with significant privacy concerns, including exposed chat logs. But Deepseek’s free offering was clearly pressing OpenAI and Microsoft to respond fast, given that R1 was wooing developers worldwide.

Tying o1 Exclusivity to Copilot​

By bundling o1's reasoning model with Microsoft Copilot instead of keeping it locked within OpenAI products, Microsoft scores a multi-pronged win:

• Increased User Adoption: Thousands, if not millions, dabbling in Copilot for the first time will suddenly realize how indispensable deep reasoning can be.
• Locking Competitors Out: Competitors like Deepseek don’t yet have a mainstream productivity tool like Microsoft Office backing their model.
• Building Faith in the Ecosystem: Even non-paying users can now taste “premium” AI capability, boosting trust and interest in Microsoft’s AI ecosystem.

---

How to Enable and Use “Think Deeper” on Copilot​

If you’re giddy with excitement and checking your apps right this moment, here’s the practical breakdown:

• Locate the "Think Deeper" Button: While using Copilot, either in a standalone chat-like interface or embedded alongside apps like Word or Excel, you should soon see a “Think Deeper” toggle or button. Note that this rollout might take a little longer to reach every single user.
• Enter Your Prompt: Ask Copilot a question as usual, but click “Think Deeper.” The model will take about 30 seconds to generate more nuanced and reasoned responses.
• Review Step-by-Step Outputs: Don’t just take it at face value—o1 showcases the logical steps leading up to its conclusions. This unique traceability makes it ideal for critical analysis and educational scenarios.
• Be Wary of Peak Times: While access to o1 is now free, you’re likely bound by some limits analogous to other free Copilot services. This could mean heavy demand during peak usage hours.

---

But Wait, What’s the Catch?​

Though this free rollout is undoubtedly a boon for users, it’s not a silver bullet:

• Offline Processing Only: Since the o1 model functions offline during "Think Deeper" operations, don’t expect it to pull up real-time knowledge for you. If you’re working on something time-sensitive requiring real-world numbers or events (hello, reporters!), you’ll still need to turn elsewhere.
• No Unlimited Free Lunch: Microsoft hasn’t clarified where, when, or if throttling will occur in the free tier. Given Copilot historically limits accessibility during high-demand periods, don’t be surprised when you face a line.

---

Broader Implications: Why You Should Care Beyond Chatbots​

Artificial intelligence is increasingly blending into every nook and cranny of our digital lives—from managing email overload to writing essays or solving engineering problems. The introduction of free access to advanced reasoning models hints at where this evolution is racing towards:

• Leveling the Playing Field: Students, small business owners, and underfunded researchers now have access to tools once considered a luxury.
• The Rise of Reasoning Agents: Tools like Copilot with reasoning models are becoming pseudo-advisors. For industries relying on precision (medicine, legal, engineering), this capability is game-changing.
• Increased Rivalry = Faster Evolution: Competition between players like OpenAI and Deepseek drives prices down for consumers while speeding up innovation cycles. Win-win.

---

Closing Thoughts: A Smart Move with Smarter Tech​

The democratization of advanced reasoning AI—especially at zero cost—signals the start of a new phase in making powerful tools universally accessible. For Windows users who live and breathe Microsoft’s productivity suite, this integration feels like slipping on a glove. No weird learning curves, no additional software—just a much smarter Copilot.
Even if you’re just a casual AI user, the ability to see how "Think Deeper" breaks down complex reasoning can fundamentally shift how you approach work, study, and beyond.
So, what are your thoughts? Have you had a chance to try Think Deeper in Copilot yet? Share your impressions and questions on our forums—let's think deeper about this one, together!

---

Source: Lifehacker https://lifehacker.com/tech/openai-think-deeper-model-free-copilot