AI models get benchmarked on math puzzles and coding challenges

but how do they perform on real bookkeeping work?

Same transactions. Same accounts. Same prompts.

‍

We’re putting them head-to-head on tasks like:

- Transaction categorisation

- VAT & tax code assignment

- Invoice parsing

- …and more

‍

Which model will come out on top?

Stay tuned — the leaderboard is almost here.