It depends what your use-case is. At a low enough cost this would work for a project I'm doing where I really just need to be able to mostly search large docume

It depends what your use-case is. At a low enough cost this would work for a project I'm doing where I really just need to be able to mostly search large documents. 100% accuracy and a lost or hallucinated paragraph here and there wouldn't be a deal-killer, especially if the original page image is available to the user too. And additionally, this also might work if you are feeding the output into a bunch of humans to proof.