After much anticipation and hype, OpenAI unveiled GPT-5 back in August, but as tools began to integrate the latest model, there was also significant backlash from customers which prompted the company to make some changes in response. While the launch of GPT-5 didn"t go as smoothly as planned, the ChatGPT creator wants to win over users again with today"s reveal of GPT-5.1.
Just like GPT-5, the latest GPT-5.1 model is also available in two flavors: Instant and Thinking. The former is a more conventional model that is good at following instructions, but it is now designed to be more "warmer" and intelligent. Meanwhile, GPT-5.1 Thinking is an advanced reasoning model which has been revamped so that it is quicker at understanding and solving simpler tasks and more persistent when handling complex items.
By default, GPT-5.1 Instant is more playful and conversational, but OpenAI is giving personalization options in ChatGPT so that the model is more attuned to your style. It also leverages "adaptive reasoning" where it decides when to think before answering complex questions, leading to better results in evaluation metrics as well.
Meawhile, GPT-5.1 Thinking is an overall better model than its previous advanced reasoning variant, but it is also designed to be more approachable as it uses less jargon, while also being more empathetic and warmer by default.
OpenAI has noted that GPT-5.1 Auto will be responsible for deciding which model to route user prompts to. In terms of availability, the latest model is rolling out in a staggered manner to ChatGPT Pro, Plus, Go, and Business right now, followed by free customers. It will also become available on the API later this week.
Regarding deprecation of older models, OpenAI stated that:
GPT‑5 (Instant and Thinking) will remain available in ChatGPT under the legacy models dropdown for paid subscribers for three months, so people have time to compare and adapt at their own pace. The GPT‑5 sunset period does not affect the availability of other legacy models. Going forward, when we introduce new ChatGPT models, our approach is to give people ample space to evaluate what’s changed and share feedback, allowing us to continue innovating our frontier models while transitioning smoothly. Sunset periods will be communicated clearly and with plenty of advance notice.
In terms of new personalization options on ChatGPT, OpenAI is introducing Professional, Candid, and Quirky tones. These are in addition to the existing Default, Friendly, and Efficient tones, which have also received some updates. Soon, ChatGPT may also ask you if you want to update your tonal preferences if it notices you requesting responses in a different tone. Keep in mind that many of these improvements are being rolled out in a phased manner so you may not see them immediately.