Typesense searching works on a per-token basis. Tokens are separated by whitespace (or a token you've specified in the

token_separators

field). AFAIK, you can't conjoin tokens after indexing, so for this to work you'd have to index your document with the value of "3x2", but then again, it wouldn't work consistently with "3 x 2" afterwards. I'm saying consistently, because you could get results based on the

drop_tokens_threshold

parameter, but what results you get is highly dependent on your dataset