Today, OpenAI has introduced the following iterations to elevate the GPT-5 series:

GPT-5.1 Instant: The most frequently utilized model, exuding warmth, heightened intelligence, and superior adherence to instructions.

GPT-5.1 Thinking: The pinnacle of reasoning models, boasting enhanced comprehensibility, swift handling of straightforward tasks, and unwavering persistence in tackling complex challenges.

Officials have emphasized that an exemplary AI should not only exhibit intelligence but also foster enjoyable interactions. GPT-5.1 has made notable strides in both intelligence and conversational style.

Users now have the flexibility to effortlessly tailor the tone of ChatGPT. Given that individual preferences for chat styles vary widely, and even the tone of each conversation can differ, the research team has introduced more intuitive and effective control mechanisms. This enables ChatGPT to align more closely with users' desired response tones.

GPT-5.1 Instant stands as the most prevalent model within the ChatGPT suite, featuring default settings that are inherently friendly, natural, and conversational. Early testing has revealed that it frequently astonishes users with its lighthearted and engaging demeanor, all while preserving clarity and practicality.

The research team has also refined the model's ability to follow instructions, empowering it to answer real-world queries with greater reliability.

GPT-5.1 Instant debuts adaptive reasoning technology, enabling it to autonomously determine when to engage in pre-answer contemplation for more challenging questions. This results in more comprehensive and accurate responses, all while maintaining swift responsiveness.

GPT-5.1 can finely tune its thinking time based on the question at hand—devoting more time to intricate problems and responding more rapidly to simpler ones. In practical scenarios, this translates to more detailed answers for difficult inquiries and quicker replies for straightforward ones.

GPT-5.1's responses are also more lucid, with a reduced reliance on technical jargon and undefined terms. This renders the most potent model more accessible and user-friendly, particularly for complex tasks in the workplace and for elucidating technical concepts.

The default tone of GPT-5.1 Thinking is also imbued with warmth and empathy.

Beyond these model enhancements, the team has simplified the process of customizing ChatGPT's tone and style.

GPT-5.1 introduces preset options for users to personalize ChatGPT's response tone. These options have been meticulously crafted to mirror common usage patterns of ChatGPT. The underlying design philosophy is to enable users to swiftly and intuitively select the most fitting personalized mode for their needs.

Users can directly fine-tune various attributes of ChatGPT in personalized settings, including the conciseness, friendliness, and readability of responses, as well as the frequency of emoji usage. ChatGPT can also proactively inquire about the preferred tone or style during conversations.

The GPT-5.1 Instant and Thinking versions will be progressively rolled out starting today, with both iterations set to be introduced to the API later this week. Both versions will feature adaptive reasoning capabilities.

References:

https://openai.com/index/gpt-5-1/