OpenAI's GPT-4o is a natively multimodal model accepting text, audio, images, and video inputs, and is competitive in agentic tasks with its GPT-5.6 Sol model.
Open the full topic