Sending records in batches

Send indexing operations in batches to reduce network calls and speed up indexing. Each batch groups multiple in a single API request. For example, sending one million records individually requires one million indexing requests. Sending them in batches of 1,000 reduces this to 1,000 requests. Algolia accepts up to , but for better indexing performance, keep each batch below 10 MB. The number of records per batch depends on their average size. Two things to keep in mind:

Batching records doesn’t reduce your operations count. Algolia counts indexing operations per record, not per method call, so from a pricing perspective, batching records is the same as indexing records individually.
Be careful when approaching your plan’s record limit. If you’re close to the record limit, batch operations may fail. The error message “You have exceeded your Record quota” means the engine doesn’t know if the batch operation updates or adds records. If this happens, upgrade to a plan with a higher record limit or reduce your batch size.

Manage records with the API

To run the code examples on this page, install the latest API client. When using the saveObjects method, the API clients automatically chunk your records into batches of 1,000. With this approach, you would make 1,000 indexing requests instead of a million. Depending on your records’ sizes and your network speed, you could create bigger or smaller chunks.

List<BatchResponse> response = await client.SaveObjectsAsync(
  "INDEX_NAME",
  new List<Object>
  {
    new Dictionary<string, string> { { "objectID", "1" }, { "name", "Adam" } },
    new Dictionary<string, string> { { "objectID", "2" }, { "name", "Benoit" } },
  }
);

response, err := client.SaveObjects(
  "INDEX_NAME",
  []map[string]any{{"objectID": "1", "name": "Adam"}, {"objectID": "2", "name": "Benoit"}})
if err != nil {
  // handle the eventual error
  panic(err)
}

// response is a slice of search.BatchResponse

List<BatchResponse> response = client.saveObjects(
  "INDEX_NAME",
  Arrays.asList(
    new HashMap() {
      {
        put("objectID", "1");
        put("name", "Adam");
      }
    },
    new HashMap() {
      {
        put("objectID", "2");
        put("name", "Benoit");
      }
    }
  )
);

const response: BatchResponse[] = await client.saveObjects({
  indexName: 'cts_e2e_saveObjects_javascript',
  objects: [
    { objectID: '1', name: 'Adam' },
    { objectID: '2', name: 'Benoit' },
  ],
});

var response: List<BatchResponse> =
  client.saveObjects(
    indexName = "INDEX_NAME",
    objects =
      listOf(
        buildJsonObject {
          put("objectID", JsonPrimitive("1"))
          put("name", JsonPrimitive("Adam"))
        },
        buildJsonObject {
          put("objectID", JsonPrimitive("2"))
          put("name", JsonPrimitive("Benoit"))
        },
      ),
  )

/** @var array<BatchResponse> $response */
$response = $client->saveObjects(
    'INDEX_NAME',
    [
        ['objectID' => '1',
            'name' => 'Adam',
        ],

        ['objectID' => '2',
            'name' => 'Benoit',
        ],
    ],
);

response: list[BatchResponse] = client.save_objects(
    index_name="INDEX_NAME",
    objects=[
        {
            "objectID": "1",
            "name": "Adam",
        },
        {
            "objectID": "2",
            "name": "Benoit",
        },
    ],
)

response = client.save_objects(
  "INDEX_NAME",
  [{objectID: "1", name: "Adam"}, {objectID: "2", name: "Benoit"}]
)
# response is an array of BatchResponse objects

val response: Seq[BatchResponse] = Await.result(
  client.saveObjects(
    indexName = "INDEX_NAME",
    objects = Seq(
      JObject(List(JField("objectID", JString("1")), JField("name", JString("Adam")))),
      JObject(List(JField("objectID", JString("2")), JField("name", JString("Benoit"))))
    )
  ),
  Duration(100, "sec")
)

let response: [BatchResponse] = try await client.saveObjects(
    indexName: "INDEX_NAME",
    objects: [["objectID": "1", "name": "Adam"], ["objectID": "2", "name": "Benoit"]]
)

If you use an Algolia-managed transformation to process records, use saveObjectsWithTransformation instead of saveObjects. Like saveObjects, it uploads records in batches, but sends each batch through the configured Push to Algolia connector. For more control over the batch size or indexing action, call chunkedPush on the Ingestion client directly. For more information, see:

Manage records in the dashboard

To add, edit, or delete individual records without code, see Manage individual records in the dashboard.
To import a file, see Import from the dashboard.

Search and Discovery platform

Optimization and Personalization

AI-powered experiences

Production and scale

Sending records in batches

Manage records with the API

Manage records in the dashboard

​Manage records with the API

​Manage records in the dashboard

Manage records with the API

Manage records in the dashboard