We can understand paying for premium AI model tiers to an extent, as long as it’s an actually useful expense and you’re getting your money’s worth of it. As far as this new, ultra-expensive ChatGPT Pro tier goes, though, there’s no reason any normal person needs this.
OpenAI has just announced ChatGPT Pro, a new premium tier that goes over the ChatGPT Plus tier. This plan grants users unlimited access to the powerful o1 model, along with o1-mini, GPT-4o, and Advanced Voice capabilities.
The key feature of ChatGPT Pro is “o1 pro mode,” a version of the o1 model that leverages increased computing power to tackle highly complex problems. Internal testing by OpenAI shows the new o1 pro mode excels in areas like data science, programming, and legal analysis, producing more consistently accurate output when it has to actually think about complex answers. o1 pro mode also seemingly outperforms its predecessors in challenging benchmarks involving math, science, and coding.
This enhanced performance comes at the cost of increased processing time. If you ever use ChatGPT Pro (which you probably shouldn’t), you will notice a progress bar while the o1 pro mode generates its responses. OpenAI is also awarding ChatGPT Pro grants to medical researchers at leading US institutions. These grants aim to support breakthroughs in areas like rare diseases, aging, and cancer immunotherapy.
The company plans to expand Pro’s capabilities to handle more computationally intensive tasks and make many of its features available to subscribers of other plans. That last part might be the only good takeaway here, because the plan is expensive. The new ChatGPT Pro tier commands a whopping $200 per month if you want to use it. That’s 10 times as much as what ChatGPT Plus, and other premium chatbot subscriptions, cost. It’s a lot, and we don’t think most people will get their money’s worth. It should also be noted that the plan doesn’t include any API access, which is billed separately. And this is still an AI, which means that even the $200/month tier can still be wrong.
If you have a legit use case for this, the plan is now available. Almost no one should be using this, though.
Source: OpenAI
