OpenAI announced updates to GPT-5.6 Luna and GPT-5.6 Sol, adding new reasoning controls for ChatGPT users. Free and Go users can now use the "Think" button for deeper reasoning, while Plus and Pro users get a slider to adjust the reasoning effort per response. The update aims to improve user experience and respond to feedback.
SynthePulse Insight · AI deep reading
Version 1 · 1 source
OpenAI announces GPT-5.6 Sol is now available to all paid users, with a new reasoning effort slider; free users get unlimited Luna chat and a 'Think' button. However, user feedback shows slider levels don't match the announcement, raising consistency questions.
On August 6, 2026, OpenAI announced that GPT-5.6 Sol now offers instant and deep reasoning for Plus and Pro users, aiming to create a consistent experience. Additionally, these users get a slider to control the reasoning effort for each ChatGPT reply. OpenAI says this is easier to use and they are listening to feedback.
However, a Plus subscriber named Uluç reported seeing only three levels, not the five mentioned in the announcement, and was unsure if they were using GPT-5.6. This suggests a possible discrepancy between the actual interface and the announcement, but this feedback is only a user's personal observation and has not received an official response.
Free and Go users will get unlimited text chat with GPT-5.6 Luna starting tomorrow, along with a new 'Think' button for more reasoning on harder problems. This marks OpenAI's expansion of advanced model features to non-paying users.
This move may aim to broaden the user base, but the specific impact is unclear. Whether free users will face limitations in reasoning capabilities is not mentioned in the announcement.
OpenAI claims that in high-stakes factual evaluations covering finance, medicine, and law, the new GPT-5.6 Sol produces 68% fewer factual errors than GPT-5.5 Instant. This figure comes from the official announcement, but no evaluation details or third-party verification are provided.
This claim is a source claim, not independent confirmation. Users should treat such benchmarks with caution, as evaluation methods and samples can affect results.
The new GPT-5.6 Sol is only available in ChatGPT's chat experience, not for work scenarios. This means work scenarios may still use older versions or other models, but the announcement does not specify.
This limitation may affect users who rely on work scenarios, but the specific impact is unknown.
This report is primarily based on OpenAI's official X post, which is a primary source, but the content is self-claimed. User feedback serves as a secondary source and represents only personal opinions. The 68% factual improvement figure is not independently verified and should be treated as a source claim.
OpenAI is enhancing accessibility by unifying models and expanding free features, but the discrepancy between the reasoning slider's actual levels and the announcement, along with unverified claims of factual improvement, suggests users should pay attention to actual experience and independent evaluation.
Primary report
Primary source