OpenAI reboots ChatGPT experience with GPT-5.1 after mixed reviews of GPT-5
ChatGPT is about to get faster and chattier OpenAI are upgraded flagship model GPT-5 to GPT-5.1.
OpenAI announced two updates to the GPT-5 series: GPT-5.1 Instant and GPT-5.1 Thinking. Both models are now accessible on ChatGPT.
GPT-5.1 Instant, essentially the standard and most widely used model, is now “warmer, more intelligent, and better at following your instructions,” according to OpenAI. Meanwhile, GPT-5.1 Thinking is an advanced reasoning model that is more responsive on simple tasks and more persistent on complex tasks.
“We heard clearly from users that great AI should not only be smart, but also pleasant to talk to,” OpenAI said in a statement. blog post. “GPT-5.1 significantly improves both intelligence and communication style.”
The company added that both models give users a way to “shape the tone of ChatGPT,” allowing people to control how the chat platform responds depending on the conversation they’re having.
Both models have been rolled out to ChatGPT Pro, Plus, Go and Business users, as well as the free tier. Those on the Enterprise and Edu plans get seven days of early access to the models before GPT-5.1 becomes the standard model. OpenAI said the models can also be accessed via the API, both with custom reasoning.
OpenAI has noted that it will update GPT-5 Pro to version 5.1 soon.
Instant and Thinking models
TThe 5.1 tag reflects improvements to the base model, and OpenAI considers it part of the GPT-5 family, trained on the same stack and data as the reasoning models. The biggest difference between 5.1 and 5 is the more natural and conversational tone, said Fidji Simo, CEO of OpenAI for Applications, in a Substack after.
“Based on early testing, it often surprises people with its playfulness while still remaining clear and useful,” OpenAI said in its post.
Instant can use adaptive reasoning to decide when to think about its answers, especially when it comes to more complicated questions. OpenAI noted that it has improved the model’s instructions so that while it remains responsive, it also answers the user’s question directly.
Recent model releases, such as Baidu‘s ERNIE-4.5-VL-28B-A3B-Thinking, have performed better than GPT-5 in benchmarks such as following instructions.
GPT-5.1 The mind can determine for itself how much reasoning power to devote to a prompt. It adapts to the type and complexity of a question, so a fuller, more complex question will take longer to answer than a simple summary request.
OpenAI said evaluations have shown that GPT-5.1 Thinking spends less time and therefore uses fewer tokens for simple tasks compared to GPT-5, allowing it to outperform the base model in terms of responsiveness.
One thing companies should note is that GPT-5.1 Thinking answers “with less jargon and fewer undefined terms.” OpenAI said removing jargon responses makes thinking more accessible when it comes to explaining technical concepts.
More personalization
Another big update to ChatGPT is increased personalization. This allows users to switch their conversations between a friendly and authoritative chat platform experience.
ChatGPT already allows users to choose preset model tone options, but the new update expands these options “to better reflect the most common ways people use ChatGPT.”
OOptions include ‘standard’, ‘friendly’ (formerly ‘listener’), ‘efficient’ (formerly ‘robot’), ‘professional’, ‘candid’ and ‘quirky’. Two other personalities, ‘cynical’ and ‘nerdy’, remain unchanged.
“We think many people will find that GPT-5.1 brings IQ and EQ together better, but one standard clearly cannot meet everyone’s needs,” said Simo. “That’s why we’re also making it easier to customize ChatGPT with a range of presets to choose from. The model has the same capabilities whether you select the default option or one of these options, but the style of responses will differ: more formal or more familiar, more playful or more direct, more or less jargon or jargon. Of course, eight personalities still don’t cover the full range of human diversity, but we know from our research that many people prefer simple, guided control over too many settings or open options.”
People can also adjust how many ChatGPT uses emojis. OpenAI offers detailed controls for responses and is experimenting with the ability to make the models more concise, warmer, or more scannable.
Save a deployment
OpenAI’s GPT-5 rollout was…less than perfect. While company leaders, including CEO Sam Altman, touted the new model’s capabilities, a decision to ban older and beloved models on ChatGPT was initially met with dissatisfaction. Worse, many early adopters found that GPT-5 did not outperform older options in domains like math, science, and writing.
This led Altman to reverse some of his statements about model removal, attributing performance issues to the GPT-5 router. The router, which automatically routes queries to the most appropriate models, is not going away, because GPT-5.1 Auto will route prompts to the model type that can best answer queries.
OpenAI notes that the GPT-5 Instant, Thinking, and Pro models are still available in ChatGPT’s model drop-down list, although paid subscribers only have three months to compare these older versions with the 5.1 update. However, the sunset period for GPT-5 does not affect models like GPT-4o.
“As we introduce new ChatGPT models in the future, our approach is to give people enough space to evaluate what has changed and share feedback, allowing us to continue to innovate our frontier models while making the transition smoothly,” the company said. “Sunset periods will be communicated clearly and well in advance.”
