OpenAI has updated GPT-5.6 Sol in ChatGPT, cutting unnecessary detail and adding a reasoning slider for paying users, while free users are being moved to the smaller GPT-5.6 Luna model. The changes, published on Aug 6, 2026, apply only to ChatGPT, not to ChatGPT Work or Codex, where GPT-5.6 Sol remains unchanged.
A More Direct GPT-5.6 Sol
GPT-5.6 Sol now reduces excess formatting and trims verbose responses. For simple questions, the model delivers a direct answer with just enough context. For complex tasks, it gives a fuller response with the key recommendation easy to spot. OpenAI claims better factual accuracy for both GPT-5.6 Sol and Luna, though those figures have not been independently verified.
The update targets ChatGPT only. ChatGPT Work already had the reasoning slider, and Codex keeps GPT-5.6 Sol as it was. Paying users on ChatGPT now get a slider with five settings to adjust response effort. The slider works on web, mobile, and desktop.
Slider Settings Explained
An OpenAI developer broke down which of the five settings fits which type of task. The developer's guidance suggests users match the reasoning level to the job at hand, from quick answers to deep analysis. Matthias Bastian, writing for The Decoder, reported the details.
The slider arrives after users largely ignored the model switcher previously. Bastian noted that adding more knobs to turn doesn't make the experience better. He questioned whether these controls actually help people use AI better, pointing out that choosing the right model mattered more than choosing among five reasoning levels.
Free Users Move to Luna
GPT-5.6 Luna becomes the default for Free and Go users later this week. Unlimited text chats for free users follow next week. Free users also get a "Think" button for harder questions, but that button only lets Luna reason longer. It does not switch to a stronger model.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Limits on file uploads, image generation, and other tools remain for free users. The shift means free users lose access to OpenAI's most powerful reasoning. Luna may perform well for its size, but smaller models are generally more error-prone and less reliable, Bastian wrote.
Accuracy Gains Claimed
OpenAI reports a 62% reduction in factual errors for Luna compared to GPT-5.5 Instant, based on internal evaluations using finance, medicine, and law prompts. For Sol, the claimed reduction is 68% against the same baseline. These numbers come from OpenAI's own testing and have not been independently verified.
The previous Instant model produced a completely wrong data analysis compared to its larger reasoning counterpart. That history raises the stakes for moving free users to a smaller model. The new Think button gives Luna extra time to reason, but it cannot summon Sol's capabilities.
What Comes Next
Later this week, Luna becomes the default for Free and Go users. Next week, unlimited text chats and the Think button arrive for free users. OpenAI has release notes for these changes, and the company says the update only affects ChatGPT, not ChatGPT Work or Codex.
Bastian's analysis suggests the slider may not improve user experience, given how few people used the model switcher. He also flagged the risk of free users losing access to the most powerful reasoning. The trade-off is clear: free users get unlimited chats, but they lose the strongest model.