**OpenAI Rolls Out GPT-5.1 Models, Bringing Back User Choice and Enhanced Personalization to ChatGPT**
OpenAI has announced a significant update to its ChatGPT product line, unveiling two new models—GPT-5.1 Instant and GPT-5.1 Thinking—as part of its ongoing effort to deliver a more intuitive, customizable, and intelligent conversational AI experience. This release addresses a key request from users: the ability to personally select the model that best suits their needs, a feature that had been removed with the earlier rollout of GPT-5, which defaulted to automatic model selection.
**A Response to User Feedback**
The initial introduction of GPT-5 in the summer aimed to simplify the user experience by letting ChatGPT automatically choose the best model for each prompt. While intended to streamline interactions, the change inadvertently removed a level of control and personalization that many users valued. The GPT-5.1 update directly responds to this feedback, restoring users’ ability to pick between different model types based on their current needs—whether they require quick, conversational replies, or more thoughtful, reasoned responses.
**Introducing GPT-5.1 Instant and GPT-5.1 Thinking**
The new GPT-5.1 models are focused on making conversations with ChatGPT both “smarter” and “warmer.” OpenAI describes GPT-5.1 as their most widely used and most conversational model to date, with improvements that make it better at understanding user instructions and providing helpful, relevant answers. Early testers noted that the updated model is not only more playful and engaging but also better at following specific instructions and adapting its tone to suit the context.
A standout feature of GPT-5.1 is its adaptive reasoning capability. The model can now “think before responding” to complex prompts, devoting more processing time to challenging tasks while maintaining speedy replies for simpler questions. This results in more accurate answers for technical or complicated requests. According to OpenAI, internal benchmarks show significant improvements, with GPT-5.1 outperforming previous versions on industry-standard tests such as the AIME 2025 mathematical benchmark and the Codeforces coding benchmark.
GPT-5.1 Thinking further optimizes this process, intentionally spending more time reasoning through difficult prompts to generate even more precise and thoughtful responses. OpenAI’s internal data suggests that GPT-5.1 can adjust its “thinking time” more effectively than earlier models: it responds almost twice as fast on straightforward prompts and takes twice as long—when necessary—for more complex ones. The company claims that the new models produce clearer, less jargon-laden answers, making conversations both more accessible and more capable than ever before.
**Rollout and Availability**
The rollout of GPT-5.1 Instant and GPT-5.1 Thinking is already underway, with paid subscribers—including those on Pro, Plus, Go, and Business plans—receiving priority access. Free and non-logged-in users will gain access in subsequent days as part of a phased release