OpenAI has eliminated the daily limit on free ChatGPT conversations and has upgraded the default model to GPT-5.6 Luna.

      TL;DR: OpenAI lifts text chat restrictions for free ChatGPT users. The default model now is GPT-5.6 Luna, which has 62% fewer factual inaccuracies. Free users will have access to a Think button, while paid users can utilize a reasoning slider for Sol.

      OpenAI has announced that it will remove usage limits on text chats for free ChatGPT users. This update allows free users to send an unlimited number of text-only messages without a daily limit. However, restrictions on image generation, file uploads, and voice tools will remain in place. Additionally, the default model for all users will shift from GPT-5.5 Instant to GPT-5.6 Luna, which is the lighter version of the GPT-5.6 family introduced in July after a government review.

      The key highlight is the improvement in accuracy. OpenAI reports that Luna generates 62% fewer factual errors compared to Instant. Free users and those subscribed to the $8/month Go plan will receive a new Think button that allows the model extra time to process more complex questions. Meanwhile, Plus and Pro subscribers ($20 and $200/month) will have access to a slider on GPT-5.6 Sol, enabling them to adjust the reasoning effort from high to extra high to pro, depending on the task at hand.

      OpenAI noted that it modified Sol for everyday tasks, including web research, writing, planning, and decision-making, leading to more direct answers with less unnecessary formatting. Luna will become the default this week, while the unlimited chats, Think button, and Sol slider features will be rolled out next week. ChatGPT reached over one billion monthly users in May, and this move to eliminate limits for free users aims to continue growing that user base rather than losing it to competitors like Claude, Gemini, or other emerging free alternatives.

      The safety aspect is also notable. OpenAI faces lawsuits claiming that excessive chatbot usage has negatively impacted vulnerable individuals, including minors. Removing usage limits expands this exposure. OpenAI states it has integrated safeguards to prevent models from engaging in romantic roleplay, age-restricted challenges, and positioning themselves as substitutes for real-life relationships for users under 18. Whether these safeguards are sufficient, especially with the new capability for unrestricted text usage, remains a matter that will ultimately be addressed by the courts, rather than by the company.