OpenAI Unveils GPT-5.1

OpenAI updates GPT-5 with new variations: GPT-5.1 Instant and Thinking.

OpenAI has updated its flagship AI model, GPT-5, introducing two variations: GPT-5.1 Instant and GPT-5.1 Thinking.

GPT-5.1 in ChatGPT is rolling out to all users this week.

It’s smarter, more reliable, and a lot more conversational.https://t.co/SA1Q1GPyxV

— OpenAI (@OpenAI) November 12, 2025

The first version is “warmer, more intelligent” and better at following instructions. The second is clearer, faster at handling simple tasks, and more persistent with complex ones.

“We have clearly heard from users that a good AI should be not only smart but also pleasant to interact with. GPT-5.1 significantly improves both aspects,” the announcement states.

Additionally, OpenAI has simplified the tone settings for ChatGPT. Communication style preferences “vary from person to person and even from conversation to conversation.” More intuitive and effective controls have been introduced.

GPT-5.1 Instant

Instant is the most popular ChatGPT model. The GPT-5.1 version by default has become more conversational. It “surprises people with its playfulness while remaining clear and helpful.”

The instruction processing has been improved, enabling the neural network to follow requests more accurately and provide more stable, relevant responses.

GPT-5.1 Instant features “adaptive thinking”—it can decide when it needs to “think” before responding. LLM shows significant improvements in mathematics and programming tests AIME 2025 and Codeforces, noted OpenAI.

GPT-5.1 Thinking

The upgrade to GPT-5 Thinking has made the model more efficient and understandable in everyday use. It now allocates time for contemplation more precisely: dedicating more attention to complex tasks and responding faster to simple ones.

By default, GPT-5.1 Thinking is also “warmer and more empathetic.”

“The release is a step forward in both capabilities and usability of all models. GPT-5.1 Auto will still direct each request to the most suitable neural network,” the announcement states.

GPT-5.1 Instant and Thinking began rolling out on November 12 for Pro, Plus, Go, and Business subscribers. Access will later be opened to all users.

“If you check ChatGPT today, you might not see GPT-5.1 immediately. We plan to roll it out gradually over the next few days to ensure stable operation for everyone. We will also update GPT-5 Pro to GPT-5.1 Pro soon,” noted OpenAI.

GPT-5 (Instant and Thinking) will remain available in ChatGPT as a legacy model for paid subscribers for another three months. Users will be able to compare the performance of new and old solutions and adapt to changes.

In the future, the company will give people “enough time” to evaluate changes and provide feedback. Periods for discontinuing support for old neural networks will be announced in advance.

Tone and Style Settings

Additionally, OpenAI has simplified the tone and style settings for ChatGPT.

“People have strong and diverse preferences about how ChatGPT should respond, and setting its tone to suit you should be easy,” the company believes.

The options “Default,” “Friendly,” and “Efficient” remain, with “Professional,” “Candid,” and “Unusual” added.

The new styles are designed with how users practically interact with the model in mind.

Personality settings apply to all models. Additionally, the company is experimenting with the ability to customize the chatbot more personally, including the degree of conciseness, warmth, or readability of responses, as well as the frequency of emoji use.

“ChatGPT can proactively suggest updating these settings during conversations when it notices you asking for a specific tone or style, without needing to go into settings,” the announcement states.

The updated styles and tone options became available on November 12, with the ability for fine-tuning to appear in experimental mode by the end of the week.

Back in August, OpenAI released the flagship AI model GPT-5.