Using a free tokeniser, you can break text down into smaller units (tokens) that can be processed by a model. Compare different models and use the manual mode for more detailed research.