# Prefix Search

In MeiliSearch, the search is possible even with a single letter sent.

A prefix search is when the documents sorting starts by comparing your search query with the start of each word in your dataset.

All words beginning with this letter are added to the bucket sort and then the other ranking rules are applied.


Since there are so many possible relevant documents with only one letter, the longer the search query is the fastest it will respond.

# Example

Given a set of words in a dataset:

film cinema movies show harry potter shine musical

If the search query is s only the documents containing the following words are returned:

  • show
  • shine

but not

  • movies
  • musical

Even if, as you can observe, there is an s in movies and musical. The prefix search begins searching at the start of the word. Notice that a prefix search is only done for the last word of a query, other words must be of the same length but can contain typos.

If the query is sho, only the documents containing show will be returned. MeiliSearch also handles typos while performing the prefix search. You can read more about the typo rules on the dedicated page.