Back in August, OpenAI had introduced GPT-5, a significant frontier model update to ChatGPT. And now three months later, OpenAI has released GPT-5.1, which it describes as “now warmer, more intelligent, and better at following your instructions.” OpenAI also describes it as an “upgrade” to GPT-5 that “makes ChatGPT smarter and more enjoyable to talk to.”
Improvements in both instant and thinking skills, OpenAI claims that GPT-5.1 focusses on both communication style and intelligence. The company summarises improvements to both the instant model and the thinking model.
The GPT-5.1 Thinking and GPT-5.1 Instant are two of the new models.
According to OpenAI, the former is “warmer, more intelligent, and better at following your instructions” than its predecessor, and the latter is “now easier to understand and faster on simple tasks, and more persistent on complex ones.” In most circumstances, queries will be auto-matched to models that are most suited to answer them.
Our most-used model, GPT-5.1 Instant, is now warmer, more intelligent, and more responsive to directions.
GPT-5.1 Thinking: our sophisticated reasoning model is now easier to grasp and faster on simple tasks, but more persistent on complex ones.
Examples show that GPT-5.1 quick responses read more like natural writing and less like responses from templated data. The softer, more conversational tone is the new standard, according to OpenAI.
“More intuitive and effective controls so ChatGPT can better match the tone you want in responses” are also being added by the firm. Default, amiable, effective, professional, honest, and eccentric are examples of personality types.
Another enhancement to the immediate model in GPT-5.1 is adherence to explicit directions. OpenAI demonstrates how the model correctly responds to cues in the above scenario, even though it is instructed to only use six words. This will be a welcome improvement over one of ChatGPT’s more annoying shortcomings if it performs as promised.
GPT-5.1 On the other hand, thinking has undergone a new tuning that causes it to work more slowly on jobs that need more thought and more quickly on rapid tasks. By default, it also takes on the warmer tone.
Similar to GPT-5, the auto-routing feature will automatically determine which model is optimal for each task; alternatively, you can choose the model you wish to employ based on prompts.
Its availability, starting today, premium users of the Pro, Plus, Go, and Business plans will be able to access GPT-5.1. The company said the rollout for paid clients will continue in the coming days. After the initial release, ChatGPT users who are signed out and free will be able to view the new model.
GPT-5 will be around for at least three months for premium members. Following the introduction of GPT-5 three months ago, OpenAI encountered substantial customer outrage, despite the company’s stated plan to remove outdated models immediately.
The old GPT-5 models will be accessible in ChatGPT’s legacy models dropdown menu for three months before they vanish, while the two new models will begin to roll out to ChatGPT customers this week.
OpenAI also said that it would increase the number of personality settings available for the models’ conversational tone as part of the update. According to the company’s blog post, the entire list of possibilities now includes Default, Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical. An “experiment for new ways to fine-tune ChatGPT’s style directly from settings” will be made available to some users this week, according to a release from OpenAI.
In a post on Substack on Wednesday, Fidji Simo, CEO of Applications at ChatGPT, stated, “We’re well past the point of one-size-fits-all, with more than 800 million users.”
Even though OpenAI CEO Sam Altman made a big deal out of the announcement of GPT-5 in August, the release did not live up to expectations. Many ChatGPT users voiced dissatisfaction with OpenAI’s decision to make it the default model for ChatGPT and were underwhelmed, especially with the small improvements. A day after the release of GPT-5, OpenAI decided to reintroduce GPT-4o as a choice due to intense pressure.
After GPT-5 failed to raise the threshold sufficiently, Microsoft, OpenAI’s strategic AI partner, has become increasingly interested in Anthropic competitor models. Anthropic’s models are now powering Copilot Researcher, GitHub Copilot, Copilot Studio, and a new Office Agent that can create Word and PowerPoint documents via Microsoft’s Copilot chat interface.
GPT-5.1 is being announced just weeks after OpenAI unveiled ChatGPT Atlas, an AI-powered web browser. It has a “agent mode” that is presently only available to ChatGPT Plus and Pro subscribers, and it functions similarly to the company’s Operator tool to perform operations in the browser on behalf of users.
Separately, Apple has stated that ChatGPT via Siri with Apple Intelligence would use GPT-5 beginning with iOS 26. The transition to GPT-5.1 will most likely occur in a future software upgrade.
For more information on the GPT-5.1 and its announcement from OpenAI here.
We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.
