OpenAI lifts text limits on ChatGPT, makes GPT-5.6 Luna default

OpenAI is removing the message cap in the free ChatGPT chat this week, moving both free users and Go subscribers to the GPT-5.6 Luna model. Until now, users who sent too many text messages received a notice asking them to wait several hours before continuing. Starting next week, text conversations will run without a ceiling. The new model replaces GPT-5.5 Instant, which has powered the free tier since May. GPT-5.6 Luna is the most compact model in the new family and includes a “Think” button that lets users request extra processing time before the answer is generated; the button is also free for text chats. Limits have not disappeared entirely: file or image uploads, image generation, and the upgraded voice mode remain subject to separate caps. In other words, texting is free at OpenAI’s expense, but anything that requires bandwidth or heavier processing stays limited.

Paying users on Plus and Pro plans receive an upgrade to the flagship GPT-5.6 Sol model. The main change is better alignment for everyday conversations, with less tendency to agree blindly and more substantive corrections. A new slider lets users control answer depth, similar to the effort setting in Claude. The paid upgrades are already live, and the changes for free plans will roll out next week.

Removing text caps signals a drop in inference costs, the price AI firms pay for real-time data processing. Reports in The Information recently hinted at an efficiency gain that enables running models at lower cost. OpenAI said the update is meant to make the latest models widely accessible and improve answer reliability, without the “glass ceiling” of usage caps. That does not mean costs have vanished, but they have fallen enough for OpenAI to let anyone write as much as they want.