BlockBeats News, August 7th, OpenAI announced that ChatGPT will be updated to GPT-5.6 Sol for Plus and Pro users, making responses more focused, increasing factual accuracy, and reducing unnecessary formatting and details. The same model will now support both instant responses and deep reasoning, allowing users to adjust a slider to control the level of thought put into each response.

OpenAI mentioned that in internal evaluations focusing on finance, healthcare, and legal questions, compared to GPT-5.5 Instant, GPT-5.6 Luna saw a reduction of approximately 62% in answers containing at least one factual inaccuracy, while GPT-5.6 Sol saw a reduction of about 68%.

GPT-5.6 Luna will become the default model for Free and Go users this week. Starting next week, free users will be able to engage in unlimited text conversations and, through a new Think button, tackle questions requiring deeper reasoning. However, they will still be subject to anti-abuse mechanisms, and features like file uploads, images, and other tools will remain restricted.

This update only applies to the ChatGPT's daily conversation experience; the GPT-5.6 Sol used in Work and Codex will not be adjusted with this release.