How do I set up an Amazon S3 bucket?

Create an S3 bucket in your chosen AWS Region. Keep Block Public Access enabled and use Bucket owner enforced Object Ownership for a private bucket. Record the bucket name and region.
Configure a dedicated AWS identity for Transloadit, restricted to the destination bucket and prefix. Follow the /s3/store permission policy, including its additional permissions for tags, region discovery and encryption. Do not use root access keys.
Save an Amazon S3 Template Credential with the access key ID (key), secret access key (secret), bucket and bucket_region. Keep AWS secrets on trusted backends. For short-lived AWS credentials, follow the Robot’s temporary-credential guidance.
Reference the credential’s name in your Template and explicitly set acl: "bucket-default". Omitting acl uses the Robot’s public-read default; private also sends an ACL. The following example omits the ACL header and writes under the policy’s uploads/ prefix:

{
  "steps": {
    ":original": {
      "robot": "/upload/handle"
    },
    "exported": {
      "robot": "/s3/store",
      "use": ":original",
      "credentials": "YOUR_S3_CREDENTIALS",
      "acl": "bucket-default",
      "path": "uploads/${unique_prefix}/${file.url_name}"
    }
  }
}

Replace YOUR_S3_CREDENTIALS with the saved credential name. Saving credentials does not authorize uploads: authenticate users and authorize uploads on your backend, and make Signature Authentication mandatory for signed browser flows. Follow the credential and instruction protection guidance before exposing an upload flow.
Test the Template with a non-sensitive file, check the completed Assembly Status, and verify the object in your bucket. A returned result URL does not grant public read access. Write permission can overwrite existing keys; choose unique paths and consider S3 versioning for recovery.

For more details about Templates and security, see our Template documentation.
For more information on saving to S3, such as how to customize file names, see the 🤖/s3/store documentation.
To upload files from your own Node.js application, follow the Amazon S3 upload tutorial. The file storage tutorials also cover curl and rclone workflows.

How do I set up an Amazon S3 bucket?

More

See also: