Sam Altman and the OpenAI team are planning something. Taking to X, Altman spoke of new features for ChatGPT, but ones that will be compute-intensive, meaning high energy usage and costs to back it up.
“Over the next few weeks, we are launching some new compute-intensive offerings. Because of the associated costs, some features will initially only be available to Pro subscribers, and some new products will have additional fees,” Altman stated in an X post.
“Our intention remains to drive the cost of intelligence down as aggressively as we can and make our services widely available, and we are confident we will get there over time. But we also want to learn what’s possible when we throw a lot of compute, at today’s model costs, at interesting new ideas”.
Over the next few weeks, we are launching some new compute-intensive offerings. Because of the associated costs, some features will initially only be available to Pro subscribers, and some new products will have additional fees.Our intention remains to drive the cost of…September 21, 2025
That is a pretty vague update from the OpenAI leader. However, what it essentially means is that those on ChatGPT’s more advanced Pro plan will soon get a selection of new features.
For those not familiar, the Pro version of ChatGPT costs $200 a month. It is designed for the most active AI users, and is most notable for its inclusion of GPT-5 Pro, a more advanced version of the model designed to answer “the hardest questions”.
Based on the post, it seems that these features will eventually come to other plans when the services are made “widely available”. This has become normal with AI chatbots, with new features launching first under a more expensive subscription.
What will these features be?
So far, the details are sparse. We know it will be computationally intensive… and that’s it. This, in the world of AI, could be just about anything.
Considering GPT-5 only launched recently, and included the GPT-5 Pro feature, it is unlikely to be a more intelligent thinking model. It is also unlikely to be a coding feature, and considering ChatGPT only recently launched an agent feature, it won’t be an update there either.
Considering the point of computational power, it is likely this will mean either an upgrade to image generation (to compete with Gemini’s new Nano Banana feature), a video generator upgrade, or the launch of the long-awaited ChatGPT web browser.
The chances are this will be either the image or video generation upgrade. Both of which are needed for ChatGPT, but especially video, with OpenAI’s Sora feature falling far behind the competition.
This would also line up with the computational intensity. If OpenAI was to launch a video generator, it would be almost guaranteed to be hidden behind an expensive paywall, at least at first.
Follow Tom's Guide on Google News, or add us as a preferred source, to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button!
More from Tom's Guide
- Google Gemini Gems just got an upgrade — here are 3 I built and how to try them
- I switched from ChatGPT to Gemini for one week — and here’s why I’m going back to ChatGPT
- Gemini finally replaces Google Assistant on Google TV — but it'll cost you
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
