extract-article-as-json-bulk

Choose this endpoint when you already hold a list of up to 20 URLs and need article content for all of them — it runs them concurrently in a single call for $0.10 flat, versus paying and waiting for each URL individually with the single endpoint. Prefer it over the single endpoint whenever you have 2 or more URLs to process at once. Not appropriate for non-article pages (use sibling endpoints for videos, meta tags, links, etc.) or when you need more than 20 URLs in one call.

• One or more URLs are unreachable or return non-200 status — those items carry a per-item error field and the failure count increments
• URL list exceeds 20 items — request may be rejected or only first 20 processed
• A URL points to a non-article page (e.g. homepage, login wall) — article extraction may return empty or partial fields with an error
• Network timeout on individual URLs — affected items return an error field rather than failing the whole batch
• Malformed URL in the list — that item returns a parse/fetch error