Imagine having a conversation with an AI that feels as natural and engaging as chatting with a close friend, while also being smarter and more insightful than ever before. That’s exactly what we’re delivering with the latest upgrade to the GPT series: GPT-5.1. But here’s where it gets exciting—this isn’t just about making AI more intelligent; it’s about making it more relatable, enjoyable, and tailored to you. Let’s dive into what’s new.
GPT-5.1: Smarter, Warmer, and More Conversational
Today, we’re introducing two powerhouse models under the GPT-5.1 umbrella:
GPT-5.1 Instant: Our most widely used model, now redesigned to be warmer, more intuitive, and better at understanding your instructions. Early users have been surprised by its playful yet precise responses—it’s like chatting with someone who’s both brilliant and fun to talk to.
GPT-5.1 Thinking: Our advanced reasoning model, now faster on simple tasks and more persistent on complex ones. It’s also easier to understand, stripping away jargon to make technical concepts accessible to everyone.
But here’s where it gets controversial: While we’ve made GPT-5.1 smarter and more conversational, some might argue that adding too much personality could distract from its core purpose—delivering accurate, objective information. What do you think? Should AI strike a balance between being relatable and remaining neutral?
Why These Changes Matter
We’ve listened to your feedback, and it’s clear: AI should be more than just a tool—it should be a companion. GPT-5.1 takes a giant leap in that direction. For instance, GPT-5.1 Instant now uses adaptive reasoning, meaning it knows when to pause and think before answering complex questions. This results in more thorough, accurate responses without sacrificing speed. It’s already showing significant improvements in math and coding evaluations, like AIME 2025 and Codeforces.
GPT-5.1 Thinking, on the other hand, is now more efficient. It spends more time on difficult problems and less on simple ones, giving you thorough answers when you need them and quick replies when you don’t. Plus, its responses are clearer and more empathetic, making it ideal for both work and personal use.
Customize ChatGPT to Fit Your Style
One size doesn’t fit all when it comes to AI conversations. That’s why we’re giving you more control over ChatGPT’s tone and style. Whether you prefer a professional, candid, quirky, or even nerdy vibe, we’ve got you covered. And this is the part most people miss—you can now fine-tune specific traits, like how warm, concise, or emoji-heavy the responses are. ChatGPT can even suggest updates to your preferences mid-conversation if it notices you’re steering the tone in a certain direction.
Bold question: With so many customization options, are we risking creating AI personas that are too tailored to individual preferences, potentially limiting its ability to adapt to diverse users? Share your thoughts in the comments!
Rollout Details and What’s Next
GPT-5.1 Instant and Thinking are rolling out today, starting with paid users (Pro, Plus, Go, Business) and then expanding to free and logged-out users. Enterprise and Edu plans get a seven-day early-access window. If you don’t see it immediately, don’t worry—we’re gradually releasing it to ensure stability for everyone.
For developers, GPT-5.1 Instant and Thinking will be available in the API later this week, with adaptive reasoning built in. And for those who want to stick with GPT-5 for now, it’ll remain accessible under the legacy models dropdown for paid subscribers for the next three months.
Looking ahead, these updates are just the beginning. We’re committed to making ChatGPT feel uniquely yours—smarter, more enjoyable, and more adaptable. But we want to hear from you: What features or improvements would you like to see next? Let’s keep the conversation going!
