ChatGPT Removes Free Text Limits—but “Unlimited” Still Has Boundaries

OpenAI announced on August 6 that ChatGPT Free users will receive unlimited text chats powered by GPT-5.6 Luna. Luna becomes the default model for Free and Go accounts during the week of August 6, with unlimited text chat and a new Think button scheduled to arrive the following week. The company says the access remains subject to abuse guardrails.

“Unlimited” applies to ordinary text conversations, not every part of ChatGPT. OpenAI says limits will continue for file uploads, images, and other tools. The change therefore removes one of the most visible friction points while preserving separate capacity controls around more compute-intensive or abuse-sensitive features.

The default model becomes the product

Free users are not only receiving more messages; they are moving from GPT-5.5 Instant to GPT-5.6 Luna. OpenAI says an internal evaluation covering financial, medical, and legal prompts found responses with at least one factual error were about 62 percent less common for Luna than for GPT-5.5 Instant. That is a company-run comparison on a selected evaluation, not a guarantee that any individual answer is correct.

The Think button also changes how free users allocate effort. A harder request can receive more reasoning without requiring a plan upgrade or a model-picker lesson. The useful habit is still to match scrutiny to consequence: ask for sources, check dates and numbers, and verify medical, legal, financial, or safety-critical claims outside the chat.

Paid plans are shifting from access to control

Plus and Pro users now receive an updated GPT-5.6 Sol in ChatGPT and a slider that adjusts how much thought the model applies. OpenAI says the Chat version is distinct from the versions used in Codex and ChatGPT Work, which are unchanged by this release. That distinction matters when people compare behavior across products that share a model family name.

When baseline text becomes abundant, paid differentiation moves toward stronger models, tools, speed, and control. Users should evaluate plans against the limits they actually hit rather than assuming a paid tier is required for an uninterrupted text conversation. Teams should also document the exact product surface and model variant used for repeatable work.

Unlimited use is not the same as private use

A missing message cap can encourage longer, more habitual conversations. That makes data hygiene more—not less—important. Do not paste secrets, credentials, confidential client material, or sensitive records simply because the session can continue. Review the product’s current data controls and use temporary or organization-managed workflows when the information warrants it.

The larger change is behavioral: a free chatbot can become an always-available layer for search, planning, writing, and decisions. Access is genuinely broader, but dependence can grow faster than verification habits. Unlimited text is most valuable when users keep the model’s uncertainty visible and preserve a path back to primary sources and human judgment.