The open source tax stack.

The mandates redrawing tax operations, and where each one stands. Every entry links to the open tooling for that jurisdiction.

Governments are rewiring tax compliance around real-time, structured data — e-invoicing clearance, digital reporting, certified billing software. This is the decade's compliance story, and much of the plumbing that meets it is open: invoice format libraries, VAT validation, and reporting standards you can read, audit, and embed.

Open source fits the format and transmission layer — the mandate logic itself. It won't give you the surrounding workflow, archiving, or liability cover a vendor contracts for.

Every tax leader is being asked for an AI plan. The open-source answer has a structural advantage the vendors can't match: tools your team runs on infrastructure you control, so client data and filings never leave the building. MCP servers give AI agents governed access to tax systems; skills and local models make the workflows portable and private.

Open source fits experimentation and data-sensitive deployment. It doesn't fit teams with nobody to run it — hosted AI with a data-processing agreement may serve them better.

Tax logic in a spreadsheet is a liability; tax logic in a black-box SaaS is a dependency. The middle path is tax law as inspectable code — engines and rules-as-code frameworks where every rate, threshold, and rule is visible, versioned, and testable. Several national administrations already work this way.

Open source fits when correctness must be auditable — you can read the rule that produced the number. It doesn't fit if nobody owns keeping the rules current when the law changes.

The most visible layer, and the most uneven. The United States has the IRS's own Direct File in the open; crypto cost-basis tooling is genuinely strong everywhere; but in much of Europe consumer filing stops at closed government APIs. What exists is here — including the accounting layer underneath it.

Open source fits preparation, calculation, and crypto reporting today. Actual submission depends on the jurisdiction — where filing channels are closed, no license can open them.

Tax law as data: statute corpora, rate scrapers, microsimulation models that score reform proposals, and the benchmarks that measure whether AI actually understands any of it. This is where policy teams, researchers, and the serious end of tax AI meet.

Open source fits research and policy analysis outright — the transparency is the point. Treat community-maintained legal corpora as a starting point, not as advice.

Four questions before you rely on anything in this index — each one answerable from signals shown on every listing.