OpenAI has announced a major shift in how people interact with its AI, removing all limits on text-based conversations for every single user. Coming just as ChatGPT surpassed the milestone of one billion weekly active users, the move aims to democratize access to high-level intelligence across its entire platform. These updates center around the introduction of the GPT-5.6 Luna model, which will now serve as the default engine for both Free and Go users, effectively replacing the older GPT-5.5 system.
Along with unrestricted texting, free users will soon see a new Think button appearing in their interface. This feature allows them to toggle on higher reasoning capabilities when tackling particularly complex queries that require deeper logic. While text is becoming unlimited, OpenAI clarified that restrictions remain in place for more resource-heavy features like file uploads, voice interactions, and image generation.
Paying subscribers are not being left behind in this rollout. Those on Plus and Pro plans will receive access to the GPT-5.6 Sol model, designed specifically for efficiency in web research, planning, and decision making. Unlike the standard model, Sol provides more compact and robust responses. These premium users also get a sophisticated thinking slider, allowing them to manually calibrate exactly how much computational effort the AI spends on a specific problem depending on its difficulty.
Internal data suggests these upgrades aren’t just about quantity but quality as well. OpenAI reports that factual errors have dropped significantly with these new models, citing a 62 percent reduction in mistakes for Luna and 68 percent for Sol compared to previous versions. The Sol model is available to paid members immediately, while free users can expect their expanded access and new tools to arrive over the next two weeks.