OpenAI just dropped its new GPT-4.1 models — including mini and nano versions — and they’re faster, cheaper, and better at coding, following instructions, and handling huge amounts of information than anything the company’s released before. Available through its API, GPT-4.1 can process up to a million tokens at once — that’s about 750,000 words — and beats previous models in real-world tasks like software engineering, all while cutting costs significantly. OpenAI says this is a major step toward its goal of building autonomous coding agents that can eventually write, test, and deploy entire apps on their own.