GPT-4.5 Released: Enhanced Accuracy, Reduced Hallucinations, and Expanded Knowledge

OpenAI announced GPT-4.5 as a transitional, general-purpose model ahead of GPT-5, emphasizing more natural interactions, broader knowledge, stronger alignment, and reduced hallucinations. It supports internet access, file and image uploads, and coding on a canvas but lacks voice, video understanding, and screen sharing. Due to GPU constraints, rollout begins with Pro users across platforms, with Plus and Team access expected soon, followed by Enterprise and Education.