OpenAI has removed the daily cap on free-tier ChatGPT text conversations, the company announced Thursday. Free users can now send unlimited text-only messages, though limits still apply to image generation, file uploads, and voice features. Alongside the change, OpenAI switched the free tier’s default model to GPT-5.6 Luna. Luna is a lighter version of the GPT-5.6 family, which launched in July following regulatory approval.
Luna produces 62% fewer factual errors than GPT-5.5 Instant, the model it replaces as the default for free accounts. OpenAI attributes the accuracy gain to changes made across the broader GPT-5.6 rollout rather than to a standalone update. The company positions Luna as the fast, everyday option in the GPT-5.6 lineup. It is meant to handle routine questions without the added processing time that heavier reasoning models require.
What changes with gpt-5.6 luna and the paid tiers
Free users and $8-per-month Go subscribers gain a new Think button that gives the model extra processing time on complex questions. Subscribers on the $20 and $200 monthly plans get a different tool instead. A slider on GPT-5.6 Sol adjusts reasoning intensity between high and pro levels. OpenAI also redesigned Sol for everyday tasks such as research, writing, organization, and evaluation, aiming for shorter answers with less unnecessary formatting.
The rollout is staggered. Luna becomes the default model for free users this week, while unlimited conversations, the Think button, and the Sol slider follow next week. As a result, the full set of changes will not be live for every user at the same time.
For casual users, the practical difference is a faster, more accurate model with no message ceiling. For paying subscribers, the Sol slider gives direct control over how much reasoning effort a given prompt receives, trading speed for depth on demand.
A competitive move as ChatGPT passes a billion users
ChatGPT reached one billion monthly users in May. Removing the message cap comes as OpenAI faces growing competition from Claude, Gemini, and a wave of newer free chatbot alternatives. Widening free access without a message limit closes one of the last gaps between OpenAI’s paid and unpaid tiers.
Still, the expansion arrives alongside ongoing legal scrutiny. Several lawsuits against OpenAI allege that heavy chatbot engagement has harmed vulnerable users, including minors, and unrestricted access could increase that exposure.
OpenAI says it has built in safeguards to block romantic roleplay and age-restricted content. Those safeguards also stop the chatbot from presenting itself as a substitute for real relationships when a user is under eighteen. They apply regardless of subscription tier, so the same rules govern a free Luna conversation and a paid Sol session.
Whether those safeguards hold up once daily limits disappear for millions of additional free users is a question. It is one more likely to land in court than in a product announcement.