Substack export

Substack exposes a public JSON archive API per publication. No auth, no HTML scraping.

https://<sub>.substack.com/api/v1/archive?sort=new&limit=50&offset=N   # metadata, paginated
https://<sub>.substack.com/api/v1/posts/<slug>                          # full body_html

Run it

python3 substack_export.py --sub <subdomain> [--sub <another>] [--out substack_posts.json]

Per post: {sub, title, subtitle, date, slug, wordcount, id, url, audience, body}, sorted oldest -> newest, bodies stripped to plain text.

Traps

• The archive caps a page at ~24 rows whatever limit asks for. A loop that stops when a page is shorter than limit sees only the first page (e.g. 23 of 51 posts). The only valid stop is an EMPTY page, and the offset advances by rows returned, never by the limit asked. The script does this.
• Only currently-public posts are returned. Posts you archived or unpublished don't appear; for a full history you need your own export (Settings -> Exports) or another copy.
• Preformatted blocks carry hidden editor boilerplate ("Text within this block will maintain its original spacing when published") inside a hide-text label. The stripper drops it; a naive tag-strip turns it into fake teaser text.
• Paywalled posts return only the free preview in body_html unless you're authenticated.