Skip to main content
Full-text search is the core feature of Meilisearch. Once you have documents in an index, you can search them with a simple query and get relevant results in milliseconds. If you haven’t added documents yet, follow the indexing getting started guide first. Send a search request to your index with the q parameter: 
curl -X POST 'MEILISEARCH_URL/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer MEILISEARCH_KEY' \
  --data-binary '{
    "q": "galaxy"
  }'
Meilisearch returns a JSON response with matching documents: 
{
  "hits": [
    {
      "id": 24,
      "title": "Guardians of the Galaxy",
      "overview": "A group of intergalactic criminals are forced to work together...",
      "genres": ["Action", "Science Fiction"]
    },
    {
      "id": 25,
      "title": "The Hitchhiker's Guide to the Galaxy",
      "overview": "Mere seconds before the Earth is to be demolished...",
      "genres": ["Adventure", "Comedy", "Science Fiction"]
    }
  ],
  "query": "galaxy",
  "processingTimeMs": 1,
  "limit": 20,
  "offset": 0,
  "estimatedTotalHits": 2
}

​
Understanding the response

FieldDescription
hitsArray of matching documents, ordered by relevance
queryThe search query you sent
processingTimeMsHow long the search took in milliseconds
limitMaximum number of results returned (default: 20)
offsetNumber of results skipped (for pagination)
estimatedTotalHitsEstimated total number of matching documents

​
Search with typos

Meilisearch handles typos automatically. A search for “galxy” or “galaxi” still returns results for “galaxy”: 
curl -X POST 'MEILISEARCH_URL/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer MEILISEARCH_KEY' \
  --data-binary '{
    "q": "galxy"
  }'
This works because Meilisearch uses typo tolerance to match words even when they contain spelling mistakes.

​
Search with multiple words

When you search with multiple words, Meilisearch finds documents containing any of those words and ranks them by how many words match: 
curl -X POST 'MEILISEARCH_URL/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer MEILISEARCH_KEY' \
  --data-binary '{
    "q": "dark knight"
  }'
Documents containing both “dark” and “knight” rank higher than documents containing only one of those words. You can control this behavior with the matching strategy.

​
Limit and paginate results

Control how many results you get back with limit and offset: 
curl -X POST 'MEILISEARCH_URL/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer MEILISEARCH_KEY' \
  --data-binary '{
    "q": "action",
    "limit": 5,
    "offset": 10
  }'
This returns 5 results starting from the 11th match.

​
Next steps

Highlight results

Show users where their query matched in each result

Phrase search

Search for exact phrases with quotes

Add filters

Narrow results with filters and sorting

Relevancy

Understand and customize how results are ranked