Microsoft Syntex AMA

This article gives a good overview of the different model types - https://learn.microsoft.com/en-gb/microsoft-365/contentunderstanding/difference-between-document-understanding-and-form-processing-model Both Unstructured and Freeform models can be used to tackle similar use cases. Limitations on file format and language may push you in one direction or other. But assuming both are possibilities for your use case, then I would start with Freeform. It has a lower bar for effort during training. If that doesn't get at the right data, then shift to Unstructured which has more of a teaching element and more need for human training. Both model types use the same OCR engine, so should be broadly the same for skewed docs. One thing to note though is that Freeform does factor in layout to the model, where Unstructured has to restructure into linear text.