A GitHub project describes classifying pages of tax documents and reports results on 261 IRS forms, at an estimated cost of about US$0.001 per page.
Published on September 18, 2026, the GitHub project presents an open-source classifier for tax-document pages based on Jev decisions. Its description reports 100% strict accuracy on 261 IRS forms and a cost of about US$0.001 per page. These are results reported by the project; the supplied material does not describe independent validation.
Engineers can consult the description and code on GitHub to examine the approach and check how the results were obtained before assessing its use. Because the material concerns tax documents, do not send real data to AI tools without authorization; prefer synthetic examples or remove identifiers, and check the data-handling policy.