I'm no expert, but my understanding is that the quality of the tool depends a lot on the specific application, and that much domain-specific customization and post-processing can be necessary to do OCR well. So you may get better advice if you share a bit about what you're trying to do with it.