OpenAI has announced the release of upgraded models, GPT-5.1 Instant and GPT-5.1 Thinking, building on their existing GPT-5 technology. These models are designed to enhance personalized conversation tones to suit user preferences better.
GPT-5.1 Instant integrates an “adaptive reasoning” feature, allowing it to determine whether a question necessitates deeper thought, thus maintaining accuracy while facilitating faster responses. This advancement has been demonstrated through significant performance improvements in math and coding benchmarks like AIME 2025 and Codeforces.
The GPT-5.1 Thinking model offers more precise control over response time, allocating more time to complex queries and responding quickly to simple ones. This results in thorough and detailed answers for difficult requests and reduced waiting times for straightforward queries. OpenAI highlights that the responses from GPT-5.1 Thinking are simplified, with fewer technical jargon and undefined expressions, making it easier for users to understand, especially in complex tasks or when explaining technical concepts.
The GPT-5.1 Auto model can automatically choose the most suitable model—either Instant or Thinking—based on user requests, ensuring optimal results without manual selection. Starting with paid users, the new models will gradually roll out to free and logged-out users. Enterprise and education plans offer a seven-day pre-access toggle before GPT-5.1 becomes the default model.
Additionally, GPT-5 Pro will be updated to GPT-5.1 Pro, and API integration is expected within the week. Both GPT-5.1 Instant (released as gpt-5.1-chat-latest) and GPT-5.1 Thinking include adaptive reasoning capabilities.
OpenAI has also made it easier to personalize ChatGPT’s tone and style, retaining previous options and adding three new ones: “professional,” “blunt and straightforward,” and “quirky and fun.” This allows users to customize interactions according to their preferences.