Migrating from Advanced Media Offloader
WP Media Cloud can take over media that Advanced Media Offloader already put in your bucket. Nothing is uploaded again and nothing in the bucket is moved, renamed or deleted.
How the migration works#
Advanced Media Offloader marks each offloaded attachment in its metadata (advmo_offloaded and advmo_path). The migration reads those marks and creates a WP Media Cloud record for each attachment: the bucket, the region and the path of the original file and every image size. WP Media Cloud then serves those files from the bucket like any file it offloaded itself.
The migration finds Advanced Media Offloader by its data, not by whether the plugin is active. You can run it while Advanced Media Offloader is active or after you deactivated it.
There are two ways to run it:
- Setup wizard (new sites): the wizard copies Advanced Media Offloader’s storage settings, saves them and imports the records in one go.
- Migration tab (sites already set up): imports the records only.
Supported providers#
| Advanced Media Offloader provider | WP Media Cloud provider | Notes |
|---|---|---|
| Cloudflare R2 | Cloudflare R2 | The account ID comes from the endpoint. |
| Amazon S3 | Amazon S3 | |
| DigitalOcean Spaces | DigitalOcean Spaces | The region comes from the endpoint. |
| Backblaze B2 | Backblaze B2 | The region comes from the endpoint. |
| Wasabi | Wasabi | |
| MinIO | Custom S3 | Endpoint and URL style are copied. |
What is imported#
- Settings (setup wizard only): provider, bucket, region, endpoint, access key and secret, and the domain as the CDN URL.
- Where settings come from: Advanced Media Offloader’s settings page, or its constants in
wp-config.php, which win when both exist. The constants are named for the provider, such asADVMO_CLOUDFLARE_R2_KEY,ADVMO_AWS_BUCKETorADVMO_WASABI_DOMAIN. - Records: one per offloaded attachment, with the paths of the original file and every size. Each record gets the bucket and region from the settings or constants. Records for attachments WP Media Cloud already tracks are skipped.
Keys stay on the server. The wizard shows Copied from Advanced Media Offloader in the key fields; the real values are never sent to your browser.
If the bucket name can’t be found in the settings or constants, the import stops with a message and creates no records. Add the bucket in Advanced Media Offloader or its constants and run it again.
Before you start#
- Back up your database.
- Install and activate WP Media Cloud. You don’t need to deactivate Advanced Media Offloader first.
- On a multisite network, run the migration on each subsite. Keys set as constants are shared by every subsite, so only a super admin can copy them; a subsite admin enters the keys by hand.
Run the migration#
New site: setup wizard#

- Open Media Cloud. The setup wizard starts.
- Under Moving from another offload plugin? you see Advanced Media Offloader with the number of files, the bucket and the provider. Click Use these settings.
- On Configure, check the fields. A box at the top lists what was copied and anything still needed. For Cloudflare R2, check the CDN URL: it must be the bucket’s public r2.dev address or a custom domain on the bucket.
- Click Continue, then Save & import N files. A progress bar shows the import.
- Follow the Next steps on the last screen, then click Go to Dashboard.
Site already set up: Migration tab#

- Set up WP Media Cloud with the same bucket Advanced Media Offloader used.
- Open Media Cloud > Migration. The Advanced Media Offloader card shows Active or Deactivated, data found, the provider, bucket, domain and the number of records.
- Click Import N Records. The import runs in batches with a progress bar until all records are imported.
- For Cloudflare R2, the card can also copy the credentials into WP Media Cloud. Advanced Media Offloader only stores S3 keys, so enter a Cloudflare API token with R2 read permission (or your account email and Global API Key) and click Copy Credentials to WPMC. The token lets WP Media Cloud list buckets and show usage stats.
After the import#
- Deactivate Advanced Media Offloader in Plugins, so only WP Media Cloud rewrites media URLs.
- Open a few pages and posts with images and check that they load.
- Open Media > Library in list view. Imported files show In cloud.
- Run Offload All Media on the Dashboard if Advanced Media Offloader had not offloaded everything. Imported files are skipped.
WP Media Cloud never changes or deletes Advanced Media Offloader’s metadata.
What happens to your URLs#
- WP Media Cloud builds each file’s URL from your CDN URL (or the bucket address) and the file’s path, so existing images keep working.
- Image URLs in post content and in
srcsetare rewritten as pages load. Nothing in the database is changed. - If Advanced Media Offloader used a custom domain, keep that domain pointing at the bucket and use it as the CDN URL.
- Versions before 1.2.3 saved imported Advanced Media Offloader records as Bunny.net. 1.2.3 corrects them once after the update, and again when you save settings, for records in the bucket WP Media Cloud is set up with.
Restoring files#
To bring every offloaded file back to your server, click Download All to Server on the Dashboard or run wp wpmc restore. This works for imported files too.
Troubleshooting#
| Problem | Fix |
|---|---|
| The wizard does not list Advanced Media Offloader | No attachment is marked as offloaded by it. |
| The import says the bucket name was not found | Add the bucket in Advanced Media Offloader’s settings or its ADVMO_ constant, then import again. |
| Keys are listed as still needed | The keys were not in the settings or constants, or you are a subsite admin on a network. Enter the keys. |
| Cloudflare R2: images are broken after the import | The CDN URL must be the bucket’s public r2.dev address or a custom domain on the bucket. Check it on the Settings tab. |
| Remove Local Copies, Clean Up or URL Replace skip imported files | Records imported before 1.2.3 may still carry the wrong provider. Save the settings once to correct them. |
| The import stops with an error | Run it again. It continues where it stopped and skips imported records. |