Inference

core

The moment a trained model actually runs — taking your input and producing an output. Training happens once, at enormous cost; inference happens every single time you generate. When you type a prompt and wait a few seconds for an image, that wait is inference: the model doing billions of calculations on GPUs. Inference cost is why AI platforms use token or credit pricing — every generation burns real compute. Faster models (like Gemini Flash) trade some quality for cheaper, snappier inference; bigger models cost more per run but deliver more.