OpenAI has updated ChatGPT, introducing GPT-5.6 Luna for free users and GPT-5.6 Sol for paid users, along with a new thinking intensity slider. Free users get unlimited text conversations, though with some restrictions. The new models show improved factual accuracy, with internal evaluations showing significant reductions in errors.
SynthePulse Insight · AI deep reading
Version 1 · 1 source
OpenAI sets GPT-5.6 Luna as the default model for free users and opens unlimited text conversations; paid users get a more focused and reliable GPT-5.6 Sol. A new thinking intensity slider lets users control reasoning depth, but anti-abuse mechanisms and tool limits remain.
OpenAI has updated ChatGPT, and the default model for free and Go users will be upgraded to GPT-5.6 Luna, along with unlimited text conversations. Plus and Pro users get the updated GPT-5.6 Sol starting today, with more focused answers and improved factual reliability.
For questions requiring deeper reasoning, free users can click the new 'Think' button to let the model spend more time on complex tasks. However, unlimited text conversations are still subject to anti-abuse mechanisms, and file uploads, images, and other tools retain usage limits.
OpenAI has added a new thinking intensity slider to ChatGPT on web, mobile, and desktop. For everyday questions, users can choose quick answers; for planning, research, writing, coding, or tasks requiring careful judgment, they can increase thinking intensity.
For Plus and Pro users, the same GPT-5.6 Sol model handles both instant answers and deep reasoning, creating a more consistent experience. When switching from quick answers to higher thinking intensity, users should perceive the model spending more time to give more comprehensive answers, rather than suddenly switching to a different model with a different tone and style.
The new GPT-5.6 Sol will better utilize searched and retrieved source information, especially in questions involving dates, numbers, sources, rules, and assumptions, reducing factual errors. In an internal evaluation, test questions included financial, medical, and legal issues requiring specific factual support.
Compared to GPT-5.5 Instant, GPT-5.6 Luna reduces answers with at least one factual error by about 62%, and GPT-5.6 Sol by about 68%. These figures come from OpenAI's internal evaluations and have not been independently verified by third parties.
The updated GPT-5.6 Sol is specifically optimized for everyday conversations in ChatGPT, so it will only be available in the Chat experience. The GPT-5.6 Sol versions supporting ChatGPT Work and Codex will not change with this release.
OpenAI states that 1 billion people use ChatGPT weekly for various tasks. This update brings ChatGPT's instant answers and deep thinking experiences closer together, ensuring the model maintains a relatively consistent tone and behavior whether in quick mode or with increased thinking intensity.
This report is based on InfoQ's retelling; all product updates and evaluation data come from OpenAI's official statements and have not been independently verified. The 62% and 68% figures from internal evaluations are self-reported by OpenAI and should be considered source claims.
OpenAI balances experience and cost between free and paid users through model tiering and a thinking intensity slider, while emphasizing improved factual reliability, but actual results still need user verification.
Primary report
Primary source