Search

Meilisearch exposes two routes to perform searches:

• A POST route: this is the preferred route when using API authentication, as it allows preflight request caching and better performance
• A GET route: the usage of this route is discouraged, unless you have good reason to do otherwise (specific caching abilities for example)

You may find exhaustive descriptions of the parameters accepted by the two routes at the end of this article.

Search for documents matching a specific query in the given index.

This is the preferred endpoint to perform search when an API key is required, as it allows for preflight requests to be cached. Caching preflight requests considerably improves search speed.

Learn more about how to use each search parameter.

By default, Meilisearch only returns an estimate of the total number of search results in a query: estimatedTotalHits. This happens because Meilisearch prioritizes relevancy and performance over providing an exhaustive number of search results. When working with estimatedTotalHits, use offset and limit to navigate between search results.

If you require the total number of search results, use the hitsPerPage and page search parameters in your query. The response to this query replaces estimatedTotalHits with totalHits and includes an extra field with number of search results pages based on your hitsPerPage: totalPages. Using totalHits and totalPages may result in slightly reduced performance, but is recommended when creating UI elements such as numbered page selectors.

Neither estimatedTotalHits nor totalHits can exceed the limit configured in the maxTotalHits index setting.

You can read more about pagination in our dedicated guide.

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "q": "american ninja" }'

{
  "hits": [
    {
      "id": 2770,
      "title": "American Pie 2",
      "poster": "https://image.tmdb.org/t/p/w1280/q4LNgUnRfltxzp3gf1MAGiK5LhV.jpg",
      "overview": "The whole gang are back and as close as ever. They decide to get even closer by spending the summer together at a beach house. They decide to hold the biggest…",
      "release_date": 997405200
    },
    {
      "id": 190859,
      "title": "American Sniper",
      "poster": "https://image.tmdb.org/t/p/w1280/svPHnYE7N5NAGO49dBmRhq0vDQ3.jpg",
      "overview": "U.S. Navy SEAL Chris Kyle takes his sole mission—protect his comrades—to heart and becomes one of the most lethal snipers in American history. His pinpoint accuracy not only saves countless lives but also makes him a prime…",
      "release_date": 1418256000
    },
    …
  ],
  "offset": 0,
  "limit": 20,
  "estimatedTotalHits": 976,
  "processingTimeMs": 35,
  "query": "american "
}

Search for documents matching a specific query in the given index.

This endpoint should only be used when no API key is required. If an API key is required, use the POST route instead.

Learn more about how to use each search parameter.

curl \
  -X GET 'http://localhost:7700/indexes/movies/search?q=american%20ninja'

{
  "hits": [
    {
      "id": 2770,
      "title": "American Pie 2",
      "poster": "https://image.tmdb.org/t/p/w1280/q4LNgUnRfltxzp3gf1MAGiK5LhV.jpg",
      "overview": "The whole gang are back and as close as ever. They decide to get even closer by spending the summer together at a beach house. They decide to hold the biggest…",
      "release_date": 997405200
    },
    {
      "id": 190859,
      "title": "American Sniper",
      "poster": "https://image.tmdb.org/t/p/w1280/svPHnYE7N5NAGO49dBmRhq0vDQ3.jpg",
      "overview": "U.S. Navy SEAL Chris Kyle takes his sole mission—protect his comrades—to heart and becomes one of the most lethal snipers in American history. His pinpoint accuracy not only saves countless lives but also makes him a prime…",
      "release_date": 1418256000
    },
    …
  ],
  "offset": 0,
  "limit": 20,
  "estimatedTotalHits": 976,
  "processingTimeMs": 35,
  "query": "american "
}

Here follows an exhaustive description of each search parameter currently available when using the search endpoint. Unless otherwise noted, all parameters are valid for the GET /indexes/{index_uid}/search, POST /indexes/{index_uid}/search, and /multi-search routes.

Parameter: q
Expected value: Any string
Default value: null

Sets the search terms.

You can search for films mentioning shifu by setting the q parameter:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "q": "shifu" }'

This will give you a list of documents that contain your query terms in at least one attribute.

{
  "hits": [
    {
      "id": 50393,
      "title": "Kung Fu Panda Holiday",
      "poster": "https://image.tmdb.org/t/p/w500/rV77WxY35LuYLOuQvBeD1nyWMuI.jpg",
      "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace.",
      "release_date": 1290729600,
      "genres": [
        "Animation",
        "Family",
        "TV Movie"
      ]
    }
  ],
  "query": "shifu"
}

Query terms go through a normalization process that removes non-spacing marks. Because of this, Meilisearch effectively ignores accents and diacritics when returning results. For example, searching for "sábia" returns documents containing "sábia", "sabiá", and "sabia".

Normalization also converts all letters to lowercase. Searching for "Video" returns the same results as searching for "video", "VIDEO", or "viDEO".

When q isn't specified, Meilisearch performs a placeholder search. A placeholder search returns all searchable documents in an index, modified by any search parameters used and sorted by that index's custom ranking rules. Since there is no query term, the built-in ranking rules do not apply.

If the index has no sort or custom ranking rules, the results are returned in the order of their internal database position.

If you enclose search terms in double quotes ("), Meilisearch will only return documents containing those terms in the order they were given. This is called a phrase search.

Phrase searches are case-insensitive and ignore soft separators such as -, ,, and :. Using a hard separator within a phrase search effectively splits it into multiple separate phrase searches: "Octavia.Butler" will return the same results as "Octavia" "Butler".

You can combine phrase search and normal queries in a single search request. In this case, Meilisearch will first fetch all documents with exact matches to the given phrase(s), and then proceed with its default behavior.

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
--data-binary '{ "q": "\"african american\" horror" }'

Use the minus (-) operator in front of a word or phrase to exclude it from search results.

The following query returns all documents that do not include the word "escape":

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "q": "-escape" }'

Negative search can be used together with phrase search:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "q": "-\"escape room\"" }'

Parameter: offset
Expected value: Any positive integer
Default value: 0

Sets the starting point in the search results, effectively skipping over a given number of documents.

Queries using offset and limit only return an estimate of the total number of search results.

You can paginate search results by making queries combining both offset and limit.

If you want to skip the first result in a query, set offset to 1:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "shifu",
    "offset": 1
  }'

Parameter: limit
Expected value: Any positive integer
Default value: 20

Sets the maximum number of documents returned by a single query.

You can paginate search results by making queries combining both offset and limit.

If you want your query to return only two documents, set limit to 2:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "shifu",
    "limit": 2
  }'

Parameter: hitsPerPage
Expected value: Any positive integer
Default value: 20

Sets the maximum number of documents returned for a single query. The value configured with this parameter dictates the number of total pages: if Meilisearch finds a total of 20 matches for a query and your hitsPerPage is set to 5, totalPages is 4.

Queries containing hitsPerPage are exhaustive and do not return an estimatedTotalHits. Instead, the response body will include totalHits and totalPages.

If you set hitsPerPage to 0, Meilisearch processes your request, but does not return any documents. In this case, the response body will include the exhaustive value for totalHits. The response body will also include totalPages, but its value will be 0.

You can use hitsPerPage and page to paginate search results.

The following example returns the first 15 results for a query:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "",
    "hitsPerPage": 15
  }'

Parameter: page
Expected value: Any positive integer
Default value: 1

Requests a specific results page. Pages are calculated using the hitsPerPage search parameter.

Queries containing page are exhaustive and do not return an estimatedTotalHits. Instead, the response body will include two new fields: totalHits and totalPages.

If you set page to 0, Meilisearch processes your request, but does not return any documents. In this case, the response body will include the exhaustive values for facetDistribution, totalPages, and totalHits.

You can use hitsPerPage and page to paginate search results.

The following example returns the second page of search results:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "",
    "page": 2
  }'

Parameter: filter
Expected value: A filter expression written as a string or an array of strings
Default value: []

Uses filter expressions to refine search results. Attributes used as filter criteria must be added to the filterableAttributes list.

For more information, read our guide on how to use filters and build filter expressions.

You can write a filter expression in string syntax using logical connectives:

"(genres = horror OR genres = mystery) AND director = 'Jordan Peele'"

You can write the same filter as an array:

[["genres = horror", "genres = mystery"], "director = 'Jordan Peele'"]

You can then use the filter in a search query:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "thriller",
    "filter": [
      [
        "genres = Horror",
        "genres = Mystery"
      ],
      "director = \"Jordan Peele\""
    ]
  }'

If your documents contain _geo data, you can use the _geoRadius and _geoBoundingBox built-in filter rules to filter results according to their geographic position.

_geoRadius(lat, lng, distance_in_meters)

curl \
  -X POST 'http://localhost:7700/indexes/restaurants/search' \
  -H 'Content-type:application/json' \
  --data-binary '{ "filter": "_geoRadius(45.472735, 9.184019, 2000)" }'

If any parameters are invalid or missing, Meilisearch returns an invalid_search_filter error.

Parameter: facets
Expected value: An array of attributes or ["*"]
Default value: null Returns the number of documents matching the current search query for each given facet. This parameter can take two values:

• An array of attributes: facets=["attributeA", "attributeB", …]
• An asterisk—this will return a count for all facets present in filterableAttributes

By default, facets returns a maximum of 100 facet values for each faceted field. You can change this value using the maxValuesPerFacet property of the faceting index settings.

When facets is set, the search results object includes the facetDistribution and facetStats fields.

facetDistribution contains the number of matching documents distributed among the values of a given facet. Each facet is represented as an object:

{
  …
 "facetDistribution": {
    "FACET_A": {
      "FACET_VALUE_X": 6,
      "FACET_VALUE_Y": 1,
    },
    "FACET_B": {
      "FACET_VALUE_Z": 3,
      "FACET_VALUE_W": 9,
    },
  },
  …
}

facetDistribution contains an object for every attribute passed to the facets parameter. Each object contains the returned values for that attribute and the count of matching documents with that value. Meilisearch does not return empty facets.

facetStats contains the lowest (min) and highest (max) numerical values across all documents in each facet. Only numeric values are considered:

{
  …
"facetStats":{
  "rating":{
    "min":2.5,
    "max":4.7
    }
  }
  …
}

If none of the matching documents have a numeric value for a facet, that facet is not included in the facetStats object. facetStats ignores string values, even if the string contains a number.

Given a movie ratings database, the following code sample returns the number of Batman movies per genre along with the minimum and maximum ratings:

curl \
  -X POST 'http://localhost:7700/indexes/movie_ratings/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "Batman",
    "facets": ["genres", "rating"]
  }'

The response shows the facet distribution for genres and rating. Since rating is a numeric field, you get its minimum and maximum values in facetStats.

{
  …
  "estimatedTotalHits":22,
  "query":"Batman",
  "facetDistribution":{
    "genres":{
      "Action":20,
      "Adventure":7,
      …
      "Thriller":3
    },
    "rating":{
      "2":1,
      …
      "9.8":1
    }
  },
  "facetStats":{
    "rating":{
      "min":2.0,
      "max":9.8
    }
  }
}

Learn more about facet distribution in the faceted search guide.

Parameter: distinct
Expected value: An attribute present in the filterableAttributes list
Default value: null

Defines one attribute in the filterableAttributes list as a distinct attribute. Distinct attributes indicate documents sharing the same value for the specified field are equivalent and only the most relevant one should be returned in search results.

This behavior is similar to the distinctAttribute index setting, but can be configured at search time. distinctAttribute acts as a default distinct attribute value you may override with distinct.

curl \
  -X POST 'http://localhost:7700/indexes/INDEX_NAME/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "QUERY TERMS",
    "distinct": "ATTRIBUTE_A"
  }'

Parameter: attributesToRetrieve
Expected value: An array of attributes or ["*"]
Default value: ["*"]

Configures which attributes will be retrieved in the returned documents.

If no value is specified, attributesToRetrieve uses the displayedAttributes list, which by default contains all attributes found in the documents.

To get only the overview and title fields, set attributesToRetrieve to ["overview", "title"].

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "shifu",
    "attributesToRetrieve": [
      "overview",
      "title"
    ]
  }'

Parameter: attributesToCrop
Expected value: An array of attributes or ["*"]
Default value: null

Crops the selected fields in the returned results to the length indicated by the cropLength parameter. When attributesToCrop is set, each returned document contains an extra field called _formatted. This object contains the cropped version of the selected attributes.

By default, crop boundaries are marked by the ellipsis character (…). You can change this by using the cropMarker search parameter.

Optionally, you can indicate a custom crop length for any attributes given to attributesToCrop: attributesToCrop=["attributeNameA:5", "attributeNameB:9"]. If configured, these values have priority over cropLength.

Instead of supplying individual attributes, you can provide ["*"] as a wildcard: attributesToCrop=["*"]. This causes _formatted to include the cropped values of all attributes present in attributesToRetrieve.

Suppose you have a field containing the following string: Donatello is a skilled and smart turtle. Leonardo is the most skilled turtle. Raphael is the strongest turtle.

Meilisearch tries to respect sentence boundaries when cropping. For example, if your search term is Leonardo and your cropLength is 6, Meilisearch will prioritize keeping the sentence together and return: Leonardo is the most skilled turtle.

If a query contains only a single search term, Meilisearch crops around the first occurrence of that term. If you search for turtle and your cropLength is 7, Meilisearch will return the first instance of that word: Donatello is a skilled and smart turtle.

If a query contains multiple search terms, Meilisearch centers the crop around the largest number of unique matches, giving priority to terms that are closer to each other and follow the original query order. If you search for skilled turtle with a cropLength of 6, Meilisearch will return Leonardo is the most skilled turtle.

If Meilisearch does not find any query terms in a field, cropping begins at the first word in that field. If you search for Michelangelo with a cropLength of 4 and this string is present in another field, Meilisearch will return Donatello is a skilled ….

If you use shifu as a search query and set the value of the cropLength parameter to 5:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "shifu",
    "attributesToCrop": ["overview"],
    "cropLength": 5
  }'

You will get the following response with the cropped text in the _formatted object:

{
  "id": 50393,
  "title": "Kung Fu Panda Holiday",
  "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
  "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
  "release_date": 1290729600,
  "_formatted": {
    "id": 50393,
    "title": "Kung Fu Panda Holiday",
    "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
    "overview": "…this year Shifu informs Po…",
    "release_date": 1290729600
  }
}

Parameter: cropLength
Expected value: A positive integer
Default value: 10

Configures the total number of words to appear in the cropped value when using attributesToCrop. If attributesToCrop is not configured, cropLength has no effect on the returned results.

Query terms are counted as part of the cropped value length. If cropLength is set to 2 and you search for one term (for example, shifu), the cropped field will contain two words in total (for example, "…Shifu informs…").

Stop words are also counted against this number. If cropLength is set to 2 and you search for one term (for example, grinch), the cropped result may contain a stop word (for example, "…the Grinch…").

If attributesToCrop uses the attributeName:number syntax to specify a custom crop length for an attribute, that value has priority over cropLength.

Parameter: cropMarker
Expected value: A string
Default value: "…"

Sets a string to mark crop boundaries when using the attributesToCrop parameter. The crop marker will be inserted on both sides of the crop. If attributesToCrop is not configured, cropMarker has no effect on the returned search results.

If cropMarker is set to null or an empty string, no markers will be included in the returned results.

Crop markers are only added where content has been removed. For example, if the cropped text includes the first word of the field value, the crop marker will not be added to the beginning of the cropped result.

When searching for shifu, you can use cropMarker to change the default …:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "shifu",
    "cropMarker": "[…]",
    "attributesToCrop": ["overview"]
  }'

{
  "id": 50393,
  "title": "Kung Fu Panda Holiday",
  "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
  "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
  "release_date": 1290729600,
  "_formatted": {
    "id": 50393,
    "title": "Kung Fu Panda Holiday",
    "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
    "overview": "[…]But this year Shifu informs Po that as Dragon Warrior,[…]",
    "release_date": 1290729600
  }
}

Parameter: attributesToHighlight
Expected value: An array of attributes or ["*"]
Default value: null

Highlights matching query terms in the specified attributes. attributesToHighlight only works on values of the following types: string, number, array, object.

When this parameter is set, returned documents include a _formatted object containing the highlighted terms.

Instead of a list of attributes, you can use ["*"]: attributesToHighlight=["*"]. In this case, all the attributes present in attributesToRetrieve will be assigned to attributesToHighlight.

By default highlighted elements are enclosed in <em> and </em> tags. You may change this by using the highlightPreTag and highlightPostTag search parameters.

The following query highlights matches present in the overview attribute:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "winter feast",
    "attributesToHighlight": ["overview"]
  }'

The highlighted version of the text would then be found in the _formatted object included in each returned document:

{
  "id": 50393,
  "title": "Kung Fu Panda Holiday",
  "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
  "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
  "release_date": 1290729600,
  "_formatted": {
    "id": 50393,
    "title": "Kung Fu Panda Holiday",
    "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
    "overview": "The <em>Winter Feast</em> is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal <em>Winter Feast</em> at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
    "release_date": 1290729600
  }
}

Parameters: highlightPreTag and highlightPostTag
Expected value: A string
Default value: "<em>" and "</em>" respectively

highlightPreTag and highlightPostTag configure, respectively, the strings to be inserted before and after a word highlighted by attributesToHighlight. If attributesToHighlight has not been configured, highlightPreTag and highlightPostTag have no effect on the returned search results.

It is possible to use highlightPreTag and highlightPostTag to enclose terms between any string of text, not only HTML tags: "<em>", "<strong>", "*", and "__" are all equally supported values.

If highlightPreTag or highlightPostTag are set to null or an empty string, nothing will be inserted respectively at the beginning or the end of a highlighted term.

The following query encloses highlighted matches in <span> tags with a class attribute:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "winter feast",
    "attributesToHighlight": ["overview"],
    "highlightPreTag": "<span class=\"highlight\">",
    "highlightPostTag": "</span>"
  }'

You can find the highlighted query terms inside the _formatted property:

{
  "id": 50393,
  "title": "Kung Fu Panda Holiday",
  "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
  "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
  "release_date": 1290729600,
  "_formatted": {
    "id": 50393,
    "title": "Kung Fu Panda Holiday",
    "poster": "https://image.tmdb.org/t/p/w1280/gp18R42TbSUlw9VnXFqyecm52lq.jpg",
    "overview": "The <span class=\"highlight\">Winter Feast</span> is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal <span class=\"highlight\">Winter Feast</span> at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
    "release_date": 1290729600
  }
}

Parameter: showMatchesPosition
Expected value: true or false
Default value: false

Adds a _matchesPosition object to the search response that contains the location of each occurrence of queried terms across all fields. This is useful when you need more control than offered by our built-in highlighting. showMatchesPosition only works for strings, numbers, and arrays of strings and numbers.

The beginning of a matching term within a field is indicated by start, and its length by length.

If you set showMatchesPosition to true and search for winter feast:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "winter feast",
    "showMatchesPosition": true
  }'

You would get the following response with information about the matches in the _matchesPosition object. Note how Meilisearch searches for winter and feast separately because of the whitespace:

{
  "id": 50393,
  "title": "Kung Fu Panda Holiday",
  "poster": "https://image.tmdb.org/t/p/w500/rV77WxY35LuYLOuQvBeD1nyWMuI.jpg",
  "overview": "The Winter Feast is Po's favorite holiday. Every year he and his father hang decorations, cook together, and serve noodle soup to the villagers. But this year Shifu informs Po that as Dragon Warrior, it is his duty to host the formal Winter Feast at the Jade Palace. Po is caught between his obligations as the Dragon Warrior and his family traditions: between Shifu and Mr. Ping.",
  "release_date": 1290729600,
  "_matchesPosition": {
    "overview": [
      {
        "start": 4,
        "length": 6
      },
      {
        "start": 11,
        "length": 5
      },
      {
        "start": 234,
        "length": 6
      },
      {
        "start": 241,
        "length": 5
      }
    ]
  }
}

Parameter: sort
Expected value: A list of attributes written as an array or as a comma-separated string
Default value: null

Sorts search results at query time according to the specified attributes and indicated order.

Each attribute in the list must be followed by a colon (:) and the preferred sorting order: either ascending (asc) or descending (desc).

When using the POST route, sort expects an array of strings.

When using the GET route, sort expects the list as a comma-separated string.

Read more about sorting search results in our dedicated guide.

You can search for science fiction books ordered from cheapest to most expensive:

curl \
  -X POST 'http://localhost:7700/indexes/books/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "science fiction",
    "sort": ["price:asc"]
  }'

When dealing with documents containing geolocation data, you can use _geoPoint to sort results based on their distance from a specific geographic location.

_geoPoint is a sorting function that requires two floating point numbers indicating a location's latitude and longitude. You must also specify whether the sort should be ascending (asc) or descending (desc):

curl \
  -X POST 'http://localhost:7700/indexes/restaurants/search' \
  -H 'Content-type:application/json' \
  --data-binary '{ "sort": ["_geoPoint(48.8561446,2.2978204):asc"] }'

Queries using _geoPoint will always include a geoDistance field containing the distance in meters between the document location and the _geoPoint:

[
  {
    "id": 1,
    "name": "Nàpiz' Milano",
    "_geo": {
      "lat": 45.4777599,
      "lng": 9.1967508
    },
    "_geoDistance": 1532
  }
]

You can read more about location-based sorting in our dedicated guide.

Parameter: matchingStrategy
Expected value: last, all, or frequency
Default value: last

Defines the strategy used to match query terms in documents.

last returns documents containing all the query terms first. If there are not enough results containing all query terms to meet the requested limit, Meilisearch will remove one query term at a time, starting from the end of the query.

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "big fat liar",
    "matchingStrategy": "last"
  }'

With the above code sample, Meilisearch will first return documents that contain all three words. If the results don't meet the requested limit, it will also return documents containing only the first two terms, big fat, followed by documents containing only big.

all only returns documents that contain all query terms. Meilisearch will not match any more documents even if there aren't enough to meet the requested limit.

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "big fat liar",
    "matchingStrategy": "all"
  }'

The above code sample would only return documents containing all three words.

frequency returns documents containing all the query terms first. If there are not enough results containing all query terms to meet the requested limit, Meilisearch will remove one query term at a time, starting with the word that is the most frequent in the dataset. frequency effectively gives more weight to terms that appear less frequently in a set of results.

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "white shirt",
    "matchingStrategy": "frequency"
  }'

In a dataset where many documents contain the term "shirt", the above code sample would prioritize documents containing "white".

Parameter: showRankingScore
Expected value: true or false
Default value: false

Adds a global ranking score field, _rankingScore, to each document. The _rankingScore is a numeric value between 0.0 and 1.0. The higher the _rankingScore, the more relevant the document.

The sort ranking rule does not influence the _rankingScore. Instead, the document order is determined by the value of the field they are sorted on.

The code sample below returns the _rankingScore when searching for dragon in movies:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "dragon",
    "showRankingScore": true
  }'

{
  "hits": [
    {
      "id": 31072,
      "title": "Dragon",
      "overview": "In a desperate attempt to save her kingdom…",
      …
      "_rankingScore": 0.92
    },
    {
      "id": 70057,
      "title": "Dragon",
      "overview": "A sinful martial arts expert wants…",
      …
      "_rankingScore": 0.91
    },
    …
  ],
  …
}

Parameter: showRankingScoreDetails
Expected value: true or false
Default value: false

Adds a detailed global ranking score field, _rankingScoreDetails, to each document. _rankingScoreDetails is an object containing a nested object for each active ranking rule.

Each ranking rule details its score in its own object. Fields vary per ranking rule.

• order: order in which this ranking rule was applied
• score: ranking score for this rule
• matchingWords: number of words in the query that match in the document
• maxMatchingWords: maximum number of words in the query that can match in the document

• order: order in which this specific ranking rule was applied
• score: ranking score for this rule
• typoCount: number of typos corrected so that the document matches the query term
• maxTypoCount: maximum number of typos accepted

• order: order in which this ranking rule was applied
• score: ranking score for this rule

• order: order in which this ranking rule was applied
• score: ranking score for this rule
• attributeRankingOrderScore: score computed from the maximum attribute ranking order for the matching attributes
• queryWordDistanceScore: score computed from the distance between the position words in the query and the position of words in matched attributes

• order: order in which this ranking rule was applied
• score: ranking score for this rule
• matchType: either exactMatch, matchesStart, or noExactMatch:
  • exactMatch: document contains an attribute matching all query terms with no other words between them and in the order they were given
  • matchesStart: document contains an attribute with all query terms in the same order as the original query
  • noExactMatch: document contains an attribute with at least one query term matching the original query
• matchingWords: the number of exact matches in an attribute when matchType is noExactMatch
• maxMatchingWords: the maximum number of exact matches in an attribute when matchType is noExactMatch

The sort ranking rule does not appear as a single field in the score details object. Instead, each sorted attribute appears as its own field, followed by a colon (:) and the sorting direction: attribute:direction.

• order: order in which this ranking rule was applied
• value: value of the field used for sorting

• order: order in which this ranking rule was applied
• value: value of the field used for sorting
• distance: same as _geoDistance

• order: order in which this specific ranking rule was applied
• value: vector used for sorting the document
• similarity: similarity score between the target vector and the value vector. 1.0 means a perfect similarity, 0.0 a perfect dissimilarity

The code sample below returns the _rankingScoreDetail when searching for dragon in movies:

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "dragon",
    "showRankingScoreDetails": true
  }'

{
  "hits": [
    {
      "id": 31072,
      "title": "Dragon",
      "overview": "In a desperate attempt to save her kingdom…",
      …
      "_rankingScoreDetails": {
        "words": {
          "order": 0,
          "matchingWords": 4,
          "maxMatchingWords": 4,
          "score": 1.0
        },
        "typo": {
          "order": 2,
          "typoCount": 1,
          "maxTypoCount": 4,
          "score": 0.75
        },
        "name:asc": {
          "order": 1,
          "value": "Dragon"
        }
      }
    },
    …
  ],
  …
}

Parameter: rankingScoreThreshold
Expected value: A number between 0.0 and 1.0
Default value: null

Excludes results below the specified ranking score.

Excluded results do not count towards estimatedTotalHits, totalHits, and facet distribution.

The following query only returns results with a ranking score bigger than 0.2:

curl \
-X POST 'http://localhost:7700/indexes/INDEX_NAME/search' \
-H 'Content-Type: application/json' \
--data-binary '{
    "q": "badman",
    "rankingScoreThreshold": 0.2
}'

Parameter: attributesToSearchOn
Expected value: A list of searchable attributes written as an array
Default value: ["*"]

Configures a query to only look for terms in the specified attributes.

Instead of a list of attributes, you can pass a wildcard value (["*"]) and null to attributesToSearchOn. In both cases, Meilisearch will search for matches in all searchable attributes.

The order of attributes in attributesToSearchOn does not affect relevancy.

The following query returns documents whose overview includes "adventure":

curl \
  -X POST 'http://localhost:7700/indexes/movies/search' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "q": "adventure",
    "attributesToSearchOn": ["overview"]
  }'

Results would not include documents containing "adventure" in other fields such as title or genre, even if these fields were present in the searchableAttributes list.

Parameter: hybrid
Expected value: An object with two fields: embedder and semanticRatio
Default value: null

Configures Meilisearch to return search results based on a query's meaning and context.

hybrid must be an object. It accepts two fields: embedder and semanticRatio.

embedder must be a string indicating an embedder configured with the /settings endpoint. It is mandatory to specify a valid embedder when performing AI-powered searches.

semanticRatio must be a number between 0.0 and 1.0 indicating the proportion between keyword and semantic search results. 0.0 causes Meilisearch to only return keyword results. 1.0 causes Meilisearch to only return meaning-based results. Defaults to 0.5.

curl -X POST 'localhost:7700/indexes/INDEX_NAME/search' \
  -H 'content-type: application/json' \
  --data-binary '{
    "q": "kitchen utensils",
    "hybrid": {
      "semanticRatio": 0.9,
      "embedder": "EMBEDDER_NAME"
    }
  }'

Parameter: vector
Expected value: an array of numbers
Default value: null

Use a custom vector to perform a search query. Must be an array of numbers corresponding to the dimensions of the custom vector.

vector is mandatory when performing searches with userProvided embedders. You may also use vector to override an embedder's automatic vector generation.

vector dimensions must match the dimensions of the embedder.

curl -X POST 'localhost:7700/indexes/INDEX_NAME/search' \
  -H 'content-type: application/json' \
  --data-binary '{ 
    "vector": [0, 1, 2],
    "embedder": "EMBEDDER_NAME"
  }'

Parameter: retrieveVectors
Expected value: true or false
Default value: false

Return document embedding data with search results. If true, Meilisearch will display vector data in each document's _vectors field.

curl -X POST 'localhost:7700/indexes/INDEX_NAME/search' \
  -H 'content-type: application/json' \
  --data-binary '{
    "q": "kitchen utensils",
    "retrieveVectors": true,
    "hybrid": {
      "embedder": "EMBEDDER_NAME"
    }
  }'

{
  "hits": [
    {
      "id": 0,
      "title": "DOCUMENT NAME",
      "_vectors": {
        "default": {
          "embeddings": [0.1, 0.2, 0.3],
          "regenerate": true
        }
      }
      …
    },
    …
  ],
  …
}

Parameter: locales
Expected value: array of supported ISO-639 locales
Default value: []

By default, Meilisearch auto-detects the language of a query. Use this parameter to explicitly state the language of a query.

In case of a mismatch between locales and the localized attributes index setting, this parameter takes precedence.

curl \
-X POST 'http://localhost:7700/indexes/INDEX_NAME/search' \
-H 'Content-Type: application/json' \
--data-binary '{
  "q": "QUERY TEXT IN JAPANESE",
  "locales": ["jpn"]
}'

{
  "hits": [
    {
      "id": 0,
      "title": "DOCUMENT NAME",
      "overview_jp": "OVERVIEW TEXT IN JAPANESE"
    }
    …
  ],
  …
}