# Documents

A document is an object composed of one or more fields. Each field consists of an attribute and its associated value.

Documents function as containers for organizing data, and are the basic building blocks of a MeiliSearch database. To search for a document, it must first be added to an index.

# Structure

document structure

# Important terms

  • Document: an object which contains data in the form of one or more fields
  • Field: a set of two data items that are linked together: an attribute and a value
  • Attribute: the first part of a field. Acts as a name or description for its associated value.
  • Value: the second part of a field, consisting of data of any valid JSON type
  • Primary Field: A special field that is mandatory in all documents. It contains the primary key and document identifier.
  • Primary Key: the attribute of the primary field. All documents in the same index must possess the same primary key. Its associated value is the document identifier.
  • Document Identifier: the value of the primary field. Every document in a given index must have a unique identifier.

# Dataset format

You can provide your dataset in the following formats:


Documents represented as JSON objects are key-value pairs enclosed by curly brackets. As such, any rule that applies to formatting JSON objects (opens new window) also applies to formatting MeiliSearch documents. For example, an attribute must be a string, while a value must be a valid JSON data type (opens new window).

As an example, let's say you are creating an index that contains information about movies. A sample document might look like this:

  "id": "1564saqw12ss",
  "title": "Kung Fu Panda",
  "genre": "Children's Animation",
  "release-year": 2008,
  "cast": [
    { "Jack Black": "Po" },
    { "Jackie Chan": "Monkey" }

In the above example, "id", "title", "genre", "release-year", and "cast" are attributes.
Each attribute must be associated with a value, e.g. "Kung Fu Panda" is the value of "title".
At minimum, the document must contain one field with the primary key attribute and a unique document id as its value. Above, that's: "id": "1564saqw12ss".


NDJSON objects consist of individual lines where each individual line is valid JSON text and each line is delimited with a newline character. Any rules that apply to formatting NDJSON (opens new window) also apply to MeiliSearch documents.

Compared to JSON, NDJSON has better writing performance and is less CPU and memory intensive. It is easier to validate and, unlike CSV, can handle nested structures.

The above JSON document would look like this in NDJSON:

  "id": "1564saqw12ss", 
  "title": "Kung Fu Panda", 
  "genre": "Children's Animation", 
  "release-year": 2008, 
  "cast": [
    { "Jack Black": "Po" },
    { "Jackie Chan": "Monkey" }


CSV files express data as a sequence of values separated by a delimiter character. Currently, MeiliSearch only supports the comma (,) delimiter. Any rules that apply to formatting CSV (opens new window) also apply to MeiliSearch documents.

Compared to JSON, CSV has better writing performance and is less CPU and memory intensive.

The above JSON document would look like this in CSV:

  "1564saqw12ss","Kung Fu Panda","Children's Animation","2008"

Since CSV does not support arrays or nested objects, cast cannot be converted to CSV.


If you don't specify the data type for an attribute, it will default to :string.

# Limitations and requirements

Documents have a soft maximum of 1000 fields; beyond that the ranking rules may no longer be effective, leading to undefined behavior.

Additionally, every document must have at minimum one field containing the primary key and a unique id.

If you try to index a document that's incorrectly formatted, missing a primary key, or possessing the wrong primary key for a given index, it will cause an error and no documents will be added.

# Fields

A field is a set of two data items linked together: an attribute and a value. Documents are made up of fields.

An attribute functions a bit like a variable in most programming languages, i.e. it is a name that allows you to store, access, and describe some data. That data is the attribute's value.

Every field has a data type dictated by its value. Every value must be a valid JSON data type (opens new window).

Take note that, in the case of strings, a value can contain at most 65535 positions. Words exceeding the 65535 position limit will be ignored.

You can also apply ranking rules to some fields. For example, you may decide recent movies should be more relevant than older ones.

If you would like to adjust how a field gets handled by MeiliSearch, you can do so in the settings.

# Field properties

A field may also possess field properties. Field properties determine the characteristics and behavior of the data added to that field.

At this time, there are two field properties: searchable and displayed. A field can have one, both, or neither of these properties. By default, all fields in a document are both displayed and searchable.

To clarify, a field may be:

  • Searchable but not displayed
  • Displayed but not searchable
  • Both displayed and searchable (default)
  • Neither displayed nor searchable

In the latter case, the field will be completely ignored when a search is performed. However, it will still be stored in the document.

# Primary field

The primary field is a special field that must be present in all documents. Its attribute is the primary key and its value is the document id.

The primary field serves the important role of uniquely identifying each document stored in an index, ensuring that it is impossible to have two exactly identical documents present in the same index.

Therefore, every document in the same index must possess the exact same primary key associated with a unique document id as value.

# Example:

Suppose we have an index called movie that contains 200,000 documents. As shown below, each document is identified by a primary field containing the primary key movie_id and a unique value (the document id).

Aside from the primary key, documents in the same index are not required to share attributes, e.g. you could have a document in this index without the "title" attribute.

    "movie_id": "1564saqw12ss",
    "title": "Kung fu Panda",
    "runtime": 95
    "movie_id": "15k1j2kkw223s",
    "title": "Batman Begins",
    "gritty reboot": true

# Primary key

The primary key is a mandatory attribute linked to a unique value the document id. It is part of the primary field.

Each index recognizes only one primary key attribute. Once a primary key has been set for an index, it cannot be changed anymore. If no primary key is found in a document, the document will not be stored.

# Setting the primary key

There are several ways to set the primary key for an index:

# Setting the primary key on index creation

The code below creates an index called movies with reference_number as primary key:

curl \
  -X POST 'http://localhost:7700/indexes' \
  -H 'Content-Type: application/json' \
  --data-binary '{
    "uid": "movies",
    "primaryKey": "reference_number"

# Setting the primary key on document addition

The code below adds a document and sets reference_number as the index's primary key:

curl \
  -X POST 'http://localhost:7700/indexes/movies/documents?primaryKey=reference_number' \
  -H 'Content-Type: application/json' \
  --data-binary '[{
      "reference_number": 287947,
      "title": "Shazam",
      "poster": "https://image.tmdb.org/t/p/w1280/xnopI5Xtky18MPhK40cZAGAOVeV.jpg",
      "overview": "A boy is given the ability to become an adult superhero in times of need with a single magic word.",
      "release_date": "2019-03-23"

# MeiliSearch guesses your primary key

If the primary key has neither been set at index creation nor as a parameter of the add documents route, MeiliSearch will search your first document for an attribute that contains the string id in a case-insensitive manner (e.g., uid, MovieId, ID, 123id123) and set it as that index's primary key.

If no corresponding attribute is found, the index will have no known primary key, and therefore, no documents will be added.

# Missing primary key error


If you get the primary_key_inference_failed error, the primary key was not recognized. This means your primary key is wrongly formatted or absent.

Manually adding the primary key can be accomplished by using its name as a parameter for the add document route or the update index route.

# Document Id

The document id is the value associated to the primary key. It is part of the primary field, and acts as a unique identifier for each of the documents of a given index.

This unique value ensures that two documents in the same index cannot be exactly alike. If two documents in the same index have the same id, then they are treated as the same document and the more recent one will replace the older.

The document id must contain only A-Z a-z 0-9 and -_ characters.

# Example:


"id": "_Aabc012_"


"id": "@BI+* ^5h2%"

Take note that the document addition request in MeiliSearch is atomic. This means that if even a single document id is incorrectly formatted, an error will occur and none of your documents will be added.

# Upload

By default, MeiliSearch limits the size of all payloads—and therefore document uploads—to 100MB.

To upload more documents in one go, it is possible to change the payload size limit at runtime using the http-payload-size-limit option.

./meilisearch --http-payload-size-limit=1048576000

The above code sets the payload limit to 1GB, instead of the 100MB default.

MeiliSearch uses a lot of RAM when indexing documents. Be aware of your RAM availability as you increase the size of your batch as this could cause MeiliSearch to crash.

When using the route to add new documents, all documents must be sent in an array even if there is only one document.

curl -X POST `http://localhost:7700/indexes/movies/documents` \
-H 'Content-Type: application/json' \
  --data-binary '[
      "movie_id": "123sq178",
      "title": "Amelie Poulain"