OpenAI has unveiled GPT-5.1, its latest flagship model, introducing two new intelligent modes — GPT-5.1 Instant and GPT-5.1 Thinking — designed to make AI interactions faster, smarter, and more engaging.
While GPT-5.1 Thinking uses extended reasoning to handle complex queries, GPT-5.1 Instant powers most ChatGPT conversations with quicker, more natural responses. Both models can now automatically assess a question’s difficulty, dedicating more processing power to tough problems while replying instantly to simpler ones.
The update also introduces eight personality modes — Default, Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical — allowing users to customise tone and style. Fine-tuned parameters can adjust emoji use and the warmth or sharpness of replies, with changes applying instantly across all chats.
Story continues below this ad
Early access to GPT-5.1 will roll out this week for Pro, Plus, Go, and Business subscribers, followed by free users soon after. Enterprise and Education plans will have a one-week preview period, while older GPT-5 versions will remain accessible via a dropdown for three months.
According to OpenAI, GPT-5.1 shows “significant improvements” in mathematical and coding performance, excelling in AIME 2025 and Codeforces evaluations. The Thinking model simplifies complex ideas by stripping away unnecessary technical jargon. “Users wanted AI that’s not just intelligent, but also enjoyable to talk to,” said Fidji Simo, CEO of Applications, adding that the update “balances depth with personality.”
Both new models will be gradually added to the API later this week to ensure platform stability.
The release comes three months after GPT-5’s August debut, which drew criticism from users who felt the model was “colder” and less consistent than earlier versions. GPT-5.1 aims to restore confidence — and connection — in AI conversations.
Expand © IE Online Media Services Pvt Ltd