Two New Ways to Sync Scryfall Data

Today we’re excited to announce two Scryfall API changes. These updates are available right now!

Gziped JSONL Bulk Files

Our current bulk files are JSON served with streaming gzip compression. This setup has caused a few workflow headaches and confusion about file size. Due to a bunch of low-level gotchas, it’s also difficult for us to meter how popular the bulk files are.

The old JSON with gzip streaming is being deprecated. Instead, we will begin serving gzipped JSONL files:

• JSONL is the perfect format for exactly what we are doing: providing a dataset export. Each JSON object in the file is on its own line. The objects are separated by newlines (\n), there is no parent JSON array wrapped around the objects, and no commas between the objects. This makes it very easy to consume by reading it from disk line-by-line or in chunks, rather than parsing the whole file at once.
• The new files will be served as .jsonl.gz. That specifically means you will be downloading a gzipped file, not transmitting the file via gzip compression. You will need to un-gzip the file on disk once you have it. It is a single file, not a tarball (it’s just .gz not .tar.gz).
• Bulk file objects will provide both formats until July 20th, 2026, when JSONL will be the only file type offered. You can fetch both from the download_uri or jsonl_download_uri properties in the meantime. After July 20th, the only property will be jsonl_download_uri and the old files will be retired.

Manifest Method

We’re adding a new /cards/manifest method. This method returns a paginated list that contains data about which Scryfall cards exist, when they were added, and when their images were last updated.

The method returns 15,000 objects per page. It doesn’t have rich data about each card other than these metrics. You can use the manifest to figure out if your local system needs to sync a new Scryfall card or image.

Need help?

If you have questions about this update, please don’t hesitate to contact us.