OpenAI announces GPT-5.1 with smarter responses and new personality types
OpenAI has started rolling out GPT-5.1, the successor to their GPT-5 model which launched in August. The company describes GPT-5.1 as an upgrade that makes ChatGPT “smarter and more enjoyable to talk to,” after the earlier version faced mixed reception.
GPT-5.1 launches in two variants: GPT-5.1 Instant and GPT-5.1 Thinking, each suited to different kinds of queries. Instant is said to be friendlier and better at following directions, while Thinking is supposed to be clearer, faster on simple requests, and more consistent for complex requests. ChatGPT will automatically choose which model best fits user queries. The older GPT-5 models remain available for three months before they get moved into the legacy section.
Show Full Article
As part of this update, OpenAI is expanding ChatGPT’s personality presets to eight types: Default, Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical. The company is also testing a new feature that allows users to fine-tune ChatGPT’s writing and conversational style directly through settings, and some users will be able to access this starting this week. OpenAI says this change was necessary as ChatGPT has more than an 800 million user base, with many expecting control over tone and behaviour.
GPT-5 launched in August amid much fanfare but did end up disappointing a few. Early adopters alleged GPT-5 only served incremental improvements and replaced GPT-4o as the default model, which raised the ire of many. Eventually, OpenAI had to bring back GPT-4o the very next day, highlighting how sensitive the company is to user feedback.
GPT-5.1 brings features users have been asking for, promising clearer responses, better following of instructions, more transparency in its reasoning. Compared to GPT-5, this update is designed to feel more responsive and less mechanical, especially in the case of everyday conversations.
While ChatGPT is the first platform one thinks of when talking about AI, competition has been heating up. Microsoft, OpenAI’s collaborator and investor, has reportedly leaned more into Anthropic’s models after GPT-5 failed to deliver a clear performance leap. Anthropic’s Claude models now power a range of Microsoft tools, including Copilot Researcher, GitHub Copilot, Copilot Studio, and a new Office Agent capable of generating documents within Word and PowerPoint. This shift suggests that companies are focussing more on a model’s effectiveness than the market share it commands.
GPT-5.1 arrives only weeks after OpenAI introduced ChatGPT Atlas, a browser driven by AI with an ‘agent mode’ that can perform actions on behalf of users. Available currently only to Plus and Pro subscribers, Atlas shows a growing push by OpenAI to build more deep, workflow-integrated AI tools.
For the everyday user of ChatGPT, GPT-5.1 will likely feel much smoother and more natural than GPT-5, particularly in tasks that require reasoning or structured responses. If you use ChatGPT to write, explain, get help with coding, or even just to have a conversation, then it makes sense to upgrade. However, if you favour speed, creativity or stability, it may be worth waiting at least a few weeks to see how GPT-5.1 performs. Users in professional contexts should also watch the benchmarks from Anthropic and Google, since competition in the AI space is moving quickly.
