- OpenAI launches GPT-5.1 Instant and GPT-5.1 Thinking, delivering improved reasoning, conversational quality, and personalization across ChatGPT.
- GPT-5.1 Instant offers faster, warmer, and more accurate responses, boosted by adaptive reasoning and stronger performance on math and coding benchmarks.
- GPT-5.1 Thinking enhances complex problem-solving, providing quicker answers for simple tasks and more detailed reasoning for challenging queries.
OpenAI has officially released OpenAI GPT-5.1, the newest update in the GPT-5 series, featuring major improvements in reasoning, conversation quality, and personalization. The new update includes two models, GPT-5.1 Instant and GPT-5.1 Thinking, designed to make ChatGPT smarter, more flexible, and more user-centered.
Smarter, more personal ChatGPT experience
According to OpenAI, the GPT‑5.1 builds on GPT‑5 with steady but important improvements. Its aim is to boost usability and intelligence, creating smoother and more efficient conversations.
The update adds adaptive reasoning, better tone control, and improved instruction following to make responses more aware of context and aligned with users. GPT-5.1 also focuses on personalization, letting users adjust how the model communicates, including tone, style, clarity, and warmth.
GPT-5.1 Instant, faster, friendlier, and more accurate
GPT-5.1 Instant, OpenAI’s main and most popular model, has been updated to sound more natural and conversational. Its ability to follow instructions has also improved, providing more accurate and context-aware responses.
A key feature of GPT-5.1 Instant is its adaptive reasoning system, which changes how much time it spends on a question depending on its difficulty. This lets it give more detailed answers to complex queries while staying quick for everyday conversations.
Early benchmark results show clear improvements in reasoning and coding tests, such as AIME 2025 and Codeforces.
GPT-5.1 Thinking, deep reasoning for complex tasks
For users working on advanced or technical challenges, GPT-5.1 Thinking delivers deeper reasoning and greater flexibility. It adjusts its processing time based on task complexity, giving quick answers to simple prompts and detailed explanations for more complex topics.
Compared to the previous GPT-5 Thinking model, the new version is said to be twice as fast on simple tasks and able to provide deeper, more organized reasoning for complex ones. OpenAI has also improved the model’s communication style, which makes its responses clearer, more understanding, and less filled with technical terms.
New personalization and usability features
GPT‑5.1 improves ChatGPT with better personalization and usability. It adds Adaptive Model Routing, which picks either the Instant or Thinking model automatically, depending on how complex your question is.
Users can now pick from six chat styles: Default, Friendly, Efficient, Professional, Candid, and Quirky. OpenAI also kept Cynical and Nerdy for those who still like the older options.
New fine-tuning controls let users adjust conciseness, warmth, readability, and emoji use. ChatGPT also suggests tone or style changes during chats. These personalization updates apply instantly across conversations for a consistent experience.
Both GPT‑5.1 models now follow custom instructions more closely, honoring user preferences and boundaries. These upgrades make ChatGPT more flexible, efficient, and better matched to each person’s communication style.
Related
OpenAI GPT-5.1 rollout and availability
The rollout of OpenAI GPT-5.1 Instant and GPT-5.1 Thinking starts with paid ChatGPT plans, including Pro, Plus, Go, and Business subscribers. Enterprise and Education plans get a seven-day early access option (off by default), with wider availability for free and logged-out users coming soon after.
Both models will also be accessible through the OpenAI API as gpt-5.1-chat-latest and gpt-5.1, supporting adaptive reasoning capabilities. Legacy GPT-5 models will remain available to paid users for a three-month transition period.
