Create Multi-GB ZIP Archives from URLs with JavaScript

Use this workflow when the total archive size is too large for the memory and lifetime of a browser tab. Eazip Cloud fetches the remote URLs and creates the ZIP outside the visitor's device.

Before you start

• Get a public key.
• Add every source hostname to the Public App.
• Use remote URLs rather than browser File or Blob objects.
• Choose stream or stored.

Create the Cloud archive

import { createZip } from '@eazip/core';

export async function downloadLargeArchive(
  urls: string[],
  publicKey: string,
) {
  const result = await createZip({
    strategy: 'cloud',
    publicKey,
    files: urls,
    zipName: 'media-export.zip',
    mode: 'stored',
    maxZipSizeBytes: 2_000_000_000,
  });

  result.downloadAll();
}

import { useEazip, EazipTray } from '@eazip/react';

export function LargeArchiveButton({
  urls,
  publicKey,
}: {
  urls: string[];
  publicKey: string;
}) {
  const zip = useEazip();

  return (
    <>
      <button
        disabled={zip.isBusy}
        onClick={() =>
          zip.download({
            strategy: 'cloud',
            publicKey,
            files: urls,
            zipName: 'media-export.zip',
            mode: 'stored',
            maxZipSizeBytes: 2_000_000_000,
          })
        }
      >
        Create ZIP archive
      </button>

      <EazipTray />
    </>
  );
}

Use stored when the prepared archive will be downloaded more than once. Omit mode to use the SDK's stream default. The public key is safe for browser code; secret credentials are not.

Split large output

Set a practical part size for the people and systems receiving the export. The example above requests parts of approximately 2 GB:

zip.download({
  strategy: 'cloud',
  publicKey,
  files: urls,
  zipName: 'export.zip',
  maxZipSizeBytes: 2_000_000_000,
});

The result exposes one zips entry per part. A file larger than the cap remains whole in its own oversized part. See Multi-ZIP splitting.

Keep the job across a reload

@eazip/react persists a created Cloud session and reconnects when the application mounts again. Core applications must save the session handle and call resumeZip() themselves.

Plan for partial output

Large URL lists should expect an occasional unavailable source. Keep successful files unless the archive is invalid without every item:

zip.download({
  strategy: 'cloud',
  publicKey,
  files: urls,
  failOnUrlError: false,
});

The tray reports a partial result and its skipped count. Cloud does not currently return the per-file error list exposed by Local jobs.

Check policy before changing code

For jobs driven by URL count rather than total bytes, continue with Zip thousands of URLs with JavaScript. You can also read Sessions and resume or Download S3 or R2 objects as a ZIP.