There are 8.3 billion people on this planet, and roughly an eighth of them use ChatGPT. OpenAI confirmed that its infamous AI chatbot now has over one billion weekly active users, a staggering number given that ChatGPT is only three and a half years old. (For reference, that's on par with global Windows users.) Love it or hate it, there's no denying that ChatGPT—and generative AI as a whole—is firmly a part of modern life.
While there are surely many paying ChatGPT customers, the vast majority are using it for free. Though it may be exciting for paid users to try out the company's latest and greatest models, the biggest impact happens whenever OpenAI makes features available free of charge. As it happens, next week, free users are getting perhaps the biggest ChatGPT update in years.
Free ChatGPT users are getting unlimited text chats #
Starting next week, you'll no longer hit a wall when simply texting with ChatGPT. OpenAI says that both free and Go users now have unlimited text chats, which should mean you can send as many text-based messages as you want without running into cooldown periods. Of course, the emphasis here is on "texts." There are absolutely still limits on other ChatGPT features, including messages with files and images, as well as Voice mode and image generation. But for users who primarily use ChatGPT with text-based messages, this update should be more than welcome.
In addition, OpenAI is switching free and Go users to the company's new GPT-5.6 Luna model, which is designed for "cost-sensitive, high-volume workloads." As such, this model likely won't be as capable as OpenAI's more powerful models available to Pro users, but free users aren't totally left in the lurch: OpenAI says there's a new "Think" button that lets you push Luna further for more complex requests. The experience is much changed from the days when users could jump between different models for free, but perhaps the unlimited text chats will make up for it for most users.
ChatGPT Pro users are getting some new perks too #
OpenAI says that Plus and Pro users are also getting some updates. The company says that its GPT-5.6 Sol model now delivers "more focused" responses, and may issue corrections instead of simply agreeing with the user when they're wrong. It also does away with some of the formatting we've come to expect from AI: Your answer may be more succinct in paragraph form, rather than expansive with bullet points. OpenAI says that Sol makes "fewer mistakes" too, but remember: all AI models hallucinate.
There's also a new slider to push the model from "Instant" to higher effort. The idea is to give users a sliding scale of power, rather than have them switch from distinct models. The more intense you push the model, the longer the responses may take, so OpenAI positions the slider as an option for users to tailor the model to their needs: If you need a quick answer, keep the slider towards the less intense side.