OpenAI has launched its latest model, GPT-5.2, on December 11, 2025, enhancing ChatGPT with improved reasoning, reduced errors, and new workflows for spreadsheets and presentations, making it invaluable for professional knowledge work. This update introduces a “three-model” approach—GPT-5.2 Instant for quick responses, GPT-5.2 Thinking for complex tasks requiring structured reasoning, and GPT-5.2 Pro for high-quality research output. Notably, the system features an Auto mode that seamlessly switches between Instant and Thinking based on user prompts. With claims of outperforming top professionals on 70.9% of GDPval tasks, GPT-5.2 offers significantly faster outputs at a fraction of the cost. OpenAI aims to reduce hallucination rates, particularly in the Thinking model. As OpenAI competes with Google’s Gemini 3, the focus remains on delivering reliable and structured work artifacts, essential in today’s competitive AI landscape. Access varies by subscription plan, ensuring tiered availability.
Source link
