ChatGPT free users get unlimited chats, but not until next week

OpenAI is taking away the most irritating limit free ChatGPT users face: the cap on text messages. The company announced the change on August 6. Open ChatGPT today, though, and you will not find unlimited chats. They come later, and OpenAI is precise about the timing in its own announcement.

What is already live

So far only the paid side has changed. Since August 6, Plus and Pro subscribers get an updated version of GPT-5.6 Sol in chat, along with a slider that sets how much thought the model puts into an answer. Leave it low for everyday questions, push it up for planning, research or coding. Instant replies and thinking now run on the same model for Plus and Pro, so the shift in tone between the two modes is gone.

This version of Sol is limited to regular chat. ChatGPT Work and Codex are unchanged.

What is still to come

Free and Go users get the change in stages. GPT-5.6 Luna becomes the default model over the course of this week, replacing GPT-5.5 Instant. Unlimited text chats and the new Think button follow in the week of August 10, according to OpenAI. The Think button gives Luna more time on harder questions, much as the slider does for paying customers.

So if you read today that the limit is gone, check the date. Readers already know the model names from the GPT-5.6 launch in July, when Sol, Terra and Luna arrived together.

Where the limits stay

Unlimited means plain text. File uploads, images and the remaining tools keep their existing caps. OpenAI also states that the unlimited chats are subject to abuse guardrails, so anyone pushing the system hard should still expect a brake.

Fewer errors, OpenAI says

OpenAI cites figures from an internal evaluation. It tested finance, medical and legal questions that turn on hard detail. Responses containing at least one factual error were about 62 percent less common with GPT-5.6 Luna than with GPT-5.5 Instant, and about 68 percent less common with GPT-5.6 Sol. No independent check of those numbers exists so far. On pricing, OpenAI had already cut the cost of its cheapest model.

New boundaries for teenagers

OpenAI has also expanded its safeguards for users under 18. The model is trained to avoid romantic roleplay and age-restricted challenges, and not to present itself as a substitute for real relationships. Age-appropriate boundaries cover sexual content, eating disorders and body image, age-restricted goods, dangerous activities and graphic violence. The system card has the details.