GPT-5.1: A new era of AI conversations

OpenAI has officially launched the GPT-5.1 engine, a new version of its GPT-5 series that promises more advanced thinking, better instruction-following, and more natural communication. GPT-5.1 is already available to ChatGPT users and represents a significant step towards combining intelligence with usability.

The model is available in two versions. GPT-5.1 Instant is designed for everyday conversations. It responds quickly, uses a friendlier tone, and follows user instructions more closely. GPT-5.1 Thinking is designed for more complex tasks. It adjusts its thinking time based on the complexity of the question, faster for easy tasks, more in-depth for more difficult ones, and offers clearer explanations.

OpenAI has also added a router that automatically selects the appropriate model based on the question, allowing for a smooth transition between quick answers and deeper thinking.

Among the improvements we find:

• Tone settings: Users can choose from default, friendly, efficient, professional, sincere, or fun tone, and customize emoji usage and brevity.
• Better following of instructions: GPT-5.1 better respects text length limits or format requirements.
• Adaptive thinking: the model assesses the complexity of the question and adjusts the thinking time accordingly.

Initial tests show improved performance on reasoning and programming tasks, largely due to the new adaptive reasoning architecture. GPT-5.1 is available first to ChatGPT Pro, Plus, Go, Business, and Edu subscribers. Free users will receive access in the coming days. Older GPT-5 models will be available for another three months. API access will follow this weekend.

After the lukewarm reception of GPT-5, OpenAI is emphasizing better communication and adaptability with GPT-5.1. The update is also aimed at developers and enterprises, allowing for faster and more reliable responses and more customization options.