Skip to main content
The export endpoint transfers data directly from one Meilisearch instance to another over the network. Unlike dumps, which create a file on disk that you must manually move, exports push data straight to a remote instance in a single operation.

​
When to use exports

  • Environment migration: Move data from a staging instance to production (or vice versa).
  • Creating replicas: Set up a second instance with the same data for redundancy or load distribution.
  • Scaling: Transfer indexes to a larger instance when your data outgrows the current one.

​
Export data to a remote instance

Send a POST request to /export on the source instance, specifying the destination URL and (optionally) an API key: 
curl \
  -X POST 'MEILISEARCH_URL/export' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer MEILISEARCH_KEY' \
  --data-binary '{
    "url": "https://destination-instance.example.com",
    "apiKey": "destination-api-key"
  }'
Meilisearch returns a summarized task object: 
{
  "taskUid": 42,
  "indexUid": null,
  "status": "enqueued",
  "type": "export",
  "enqueuedAt": "2025-01-01T00:00:00.000000Z"
}

​
Monitor the export task

The export runs asynchronously. Use the task UID to check its progress: 
curl \
  -X GET 'MEILISEARCH_URL/tasks/42' \
  -H 'Authorization: Bearer MEILISEARCH_KEY'
When the task status changes to succeeded, all data has been transferred to the destination instance.

​
Export vs. dumps

ExportDump
MechanismDirect network transfer to a remote instanceCreates a file on the source instance’s disk
Best forLive migration between running instancesBackups, version upgrades, offline transfers
RequiresNetwork access to the destinationFile system access to move the dump file

​
Next steps

Export API reference

Full API reference for the export endpoint

Monitor tasks

Track the status of asynchronous operations

Indexing overview

Learn more about how indexing works in Meilisearch