Fluent's query builder provides a simple interface for creating complex database queries. The Query class itself (raw queries excluded) is the sole method by which Fluent communicates with your database.


You can create a new query builder from any model class.

let query = try Post.makeQuery()

You can also create queries from an instance. This is especially useful if you need to use a special database connection (like for transactions) to save or update a model.

guard let post = try Post.find(42) else { ... }
post.content = "Updated"
let query = try post.makeQuery(conn).save()


You have multiple options for fetching the results of a query.


The simplest option, .all() returns all rows relevant to the query.

let users = try User.makeQuery().filter(...).all()


You can take only the first row as well with .first().

let user = try User.makeQuery().filter(...).first()

Fluent will automatically limit the results to 1 to increase the performance of the query.


If you want to fetch a large amount of models from the database, using .chunk() can help reduce the amount of memory required for the query by fetching chunks of data at a time.

User.makeQuery().filter(...).chunk(32) { users in


Filters allow you to choose exactly what subset of data you want to modify or fetch. There are three different types of filters.


Compare filters perform a comparison between a field on your model in the database and a supplied value.

try query.filter("age", .greaterThanOrEquals, 21)

You can also use operators.

try query.filter("age" >= 21)
Case Operator Type
.equals == Equals
.greaterThan > Greater Than
.lessThan < Less Than
.greaterThanOrEquals >= Greater Than Or Equals
.lessThanOrEquals <= Less Than Or Equals
.notEquals != Not Equals
.hasSuffix Has Suffix
.hasPrefix Has Prefix
.contains Contains
.custom(String) Custom


You can omit the comparison type for .equals, e.g., query.filter("age", 23)


You can also filter by fields being in a set of data.

try query.filter("favoriteColor", in: ["pink", "blue"])

Or the opposite.

try query.filter("favoriteColor", notIn: ["brown", "black"])


By default, all query filters are joined by AND logic. You can create groups of filters within your query that are joined with AND or OR logic.

try query.or { orGroup in
    try orGroup.filter("age", .greaterThan, 75)
    try orGroup.filter("age", .lessThan, 18)

This will result in SQL similar to the following:

SELECT * FROM `users` WHERE (`age` > 75 OR `age` < 18);

.and() is also available in case you need to switch back to joining filters with AND nested inside of an OR.

Complex Example

let users = try User
    .filter("planetOfOrigin", .greaterThan, "Earth")
    .or { orGroup in
        orGroup.and { andGroup in
            andGroup.filter("name", "Rick")
            andGroup.filter("favoriteFood", "Beer")
        orGroup.and { andGroup in
            andGroup.filter("name", "Morty")
            andGroup.filter("favoriteFood", "Eyeholes")

This will result in SQL similar to the following:

SELECT * FROM `users` 
    WHERE `planetOfOrigin` = 'Earth' AND (
           (`name` = 'Rick' AND `favoriteFood` = 'Beer')
        OR (`name` = 'Morty' AND `favoriteFood` = 'Eyeholes')


Keep in mind that the AND/OR logic for a group applies only to the filters added within the group. All filters outside of a filter group will be joined by AND.


Raw filters can be used to filter by values that should not be parameterized.

try query.filter(raw: "date >= CURRENT_TIMESTAMP")


To select only distinct models from the database, add .distinct() to your query.

try query.distinct()

Limit / Offset

To limit or offset your query, use the .limit() method.

try query.limit(20, offset: 5)


To sort the results of your query, use the .sort() method

try query.sort("age", .descending)


Should you need to perform a query that the query builder does not support, you can use the raw query.

try drop.database?.raw("SELECT @@version")

You can also use the database of a given model.

User.database?.raw("SELECT * FROM `users`")

Besides providing a more expressive interface for querying your database, the query builder also takes measures to increase security by automatically sanitizing input. Because of this, try to use the query class wherever you can over performing raw queries.