I don't believe this is surprising, base58 encoding exist in Bitcoin addresses, Solana addresses, IPFS hashes and there are countless repos in github with the code. Plenty of examples in the training set. It knows base58.
Other encodings, handled without tool calls, might be more difficult unless that mapping was captured somewhere.
Well I would hope so. That's the most basic "translation" a computer can do. Converting one set of characters into another set of characters from a lookup table. How many r's in strawberry is a more difficult question.