OpenAI Removes ChatGPT Text Limits for Free Users, Defaults to GPT-5.6 Luna

OpenAI is removing text-chat limits for all users on ChatGPT, with the new GPT-5.6 Luna model becoming the default for Free and Go users, replacing GPT-5.5. The company says factual errors are roughly 62% less common on the new model. Resource-intensive features (image generation, advanced tools) will retain usage limits, but core text chat is now unlimited on the free tier.

Paid Plus and Pro users get an updated GPT-5.6 Sol powering all chats — including Instant — for one consistent experience. In OpenAI's high-stakes factuality evaluation covering finance, medicine, and law, the new GPT-5.6 Sol produced 68% fewer responses with factual errors than GPT-5.5 Instant. A new slider/Think button lets users control how long the model reasons; Simon Willison noted you can long-press the send button in the mobile app to adjust effort.

The move comes as ChatGPT crossed 1 billion weekly users, and OpenAI is positioning the change as part of a broader push to 'empower everyone with intelligence,' per president Greg Brockman, whose announcement post drew 2,400+ likes. OpenAI also spotlighted ChatGPT Work for agentic workplace productivity.

Competitively, removing free-tier limits pressures rivals like Google's Gemini and Anthropic's Claude, which still meter free usage. Some enterprise observers noted that frontier free plans now exceed most users' needs, muddying the calculus for paid upgrades. The key question is cost: serving unlimited free chats at 1B+ weekly users is an enormous inference bill, and it signals OpenAI's confidence that Luna's efficiency gains make the economics work.