Major Upgrade to ChatGPT Brings More Reliable Answers and Greater Control Over Model’s Reasoning
In a significant move, OpenAI has rolled out a major upgrade to its popular chatbot platform, ChatGPT. The update brings more reliable answers, greater control over the model’s reasoning, and improved performance for users of all levels, including those who don’t pay for a subscription.
The new version of GPT-5.6 Sol is designed specifically for everyday conversations and offers more focused answers that adapt to the complexity of the question. OpenAI claims that this updated model delivers more accurate responses and reduces factual errors by 68% compared to its predecessor, GPT-5.5 Instant. The company also notes that the new model avoids unnecessary formatting and provides helpful corrections when simply agreeing wouldn’t be useful.
One of the most significant features of the upgrade is the addition of a reasoning slider, which gives users greater control over the model’s performance. With this feature, users can adjust the level of detail and complexity of responses to suit their needs. For example, for everyday questions, users can choose the “Instant” setting, while for more complex tasks such as research or coding, they can opt for the “High” setting.
Free users will also benefit from the upgrade, with unlimited text chats using GPT-5.6 Luna. This model is designed to provide additional time for harder questions and has been shown to reduce factual errors by 62% in internal tests. OpenAI plans to remove rate limits on text chats for Free and Go users next week, although file uploads and image generation will still be subject to usage limits.
The update does not affect GPT-5.6 Sol in Codex or Work, which are designed for more advanced use cases such as research and development. Additionally, OpenAI is introducing new protections for users believed to be under 18, including stricter boundaries around sensitive topics.
The rollout of the upgrade will occur gradually, so users may not notice any changes immediately. However, with this major upgrade, ChatGPT has taken a significant step forward in its mission to provide accurate and reliable answers to users’ questions.
To take advantage of these improvements, it’s essential for users to test every layer of their ChatGPT experience before attackers do. By doing so, you can ensure that your interactions with the platform are secure and effective.
Source: Bleeping Computer — 2026-08-06