ChatGPT Free Gets Unlimited Text Chats and GPT-5.6 Luna as Default

OpenAI announced an update on August 6, 2026 that touches both the free tier and the paid plans of ChatGPT ¹. The larger change is on the free side: Free and Go switch their default model to GPT‑5.6 Luna, and text chats become unlimited ¹. OpenAI describes the update as improving everyday conversations while expanding access for Free users ¹.

What becomes unlimited is text chat specifically. Limits still apply to file uploads, images, and other tools ¹. That boundary matters when estimating how far a team can rely on the free tier for real work.

What Changes for Free, and What Changes for Paid

On Free and Go, GPT‑5.6 Luna becomes the default model this week ¹. Unlimited text chats and a “Think” button for harder questions arrive starting next week, subject to abuse guardrails ¹. The Think button gives GPT‑5.6 Luna more time to work through questions that need deeper reasoning ¹.

On Plus and Pro, the updated GPT‑5.6 Sol and a new slider are available starting the day of the announcement ¹. The slider lets you choose how much thought ChatGPT puts into each response, and works on web, mobile, and desktop ¹. OpenAI suggests keeping it quick for everyday questions and moving it up for planning, research, writing, coding, or decisions that need more thought ¹.

One structural change accompanies this on the paid side. For Plus and Pro, the same model now powers both Instant responses and deeper reasoning, which OpenAI says creates one consistent experience ¹. Moving from Instant to higher effort should feel like the model is taking extra time for a more comprehensive answer, rather than switching to a different model with its own tone or style ¹.

Because this version of GPT‑5.6 Sol is optimized for everyday chats, it is only available in the Chat experience in ChatGPT. The version of GPT‑5.6 Sol that powers Work and Codex is not changing as part of this release ¹. Anyone using Codex for development work is unaffected by this update.

What “62% Fewer Factual Errors” Actually Measures

On quality, the claim OpenAI puts forward concerns factual accuracy. In an internal evaluation of financial, medical, and legal prompts requiring factual detail, responses containing at least one factual error were about 62% less common with GPT‑5.6 Luna and 68% less common with GPT‑5.6 Sol than with GPT‑5.5 Instant ¹.

This is OpenAI’s own internal evaluation rather than third-party verification, and the comparison point is one specific model, GPT‑5.5 Instant. Even so, the choice of finance, medicine, and law — three areas where being wrong is expensive — signals which use cases the work was aimed at. OpenAI says the new GPT‑5.6 Sol is designed to make fewer mistakes when answers depend on dates, numbers, sources, rules, or assumptions, by making better use of the sources it finds ¹.

On style, the stated goals are more direct responses, tighter formatting, and avoiding extra detail where it does not help ¹. The model adapts its level of detail to the question and offers a helpful correction where simply agreeing would not be useful ¹.

Measures for Users Under 18

The release also includes measures aimed at users believed to be under 18. OpenAI says it trained the model to avoid romantic roleplay, age-restricted challenges, and presenting itself as a substitute for real-world relationships ¹.

It also applied age-appropriate boundaries around sexual content, eating disorders and body-image risks, age-restricted goods, dangerous activities, and graphic violence, and had the model encourage connection with trusted people when a teen may need support ¹. OpenAI says it reinforced this training with system-level protections and added new evaluations for how its models perform for users under 18 ¹. Details of the safety training and evaluations are in the system card ¹.

The Price-Cut Cycle Reaches the Consumer Side

GPT‑5.6 became generally available on July 9 in a three-tier Sol / Terra / Luna structure. Three weeks later, Luna was cut by 80% on the API side on July 30. OpenAI CFO Sarah Friar published an essay framing price cuts and investment as a single cycle, arguing that a falling cost of intelligence drives adoption and further investment.

This week’s free-tier expansion can be read as that same cycle surfacing in the consumer product rather than the API. OpenAI positions the update as a concrete step toward more abundant intelligence, citing wider availability of its latest models, more useful and reliable answers, and free users being able to keep text chats going without a rate limit ¹. OpenAI says 1 billion people use ChatGPT every week ¹.

For anyone weighing an AI rollout internally, this makes “hand out the free tier and see what happens” a more realistic option. But limits remain on reading documents and working with images, so unlimited text chat does not remove the case for a paid plan. The thinking slider on Plus and Pro also puts the speed-versus-depth tradeoff in the user’s hands, which means any team standardizing on ChatGPT now has one more control to explain.