Paginate a large Apify dataset and export it as CSV, JSON, or JSONL

Documented steps

1. GET https://api.apify.com/v2/datasets/{datasetId}/items — datasetId is the dataset ID or 'username~dataset-name'. A run's dataset ID is the defaultDatasetId field on its Run object.
2. Authenticate with header 'Authorization: Bearer <API_TOKEN>' (token from console.apify.com Integrations page). The docs mark the '?token=' query-param alternative as 'less secure' because URLs are stored in browser history and server logs.
3. Select the export format with 'format': json (default), jsonl, xml, html, csv, xlsx, or rss.
4. Paginate with 'offset' (default 0) and 'limit' (default unlimited — no documented maximum). Set an explicit limit such as 1000 and increment offset by the returned count each iteration.
5. Read the response headers X-Apify-Pagination-Offset, X-Apify-Pagination-Limit, X-Apify-Pagination-Count and X-Apify-Pagination-Total to drive the loop and know when to stop, rather than guessing at the total.
6. Shape the payload with 'clean' (skip empty items and '#'-prefixed fields), 'fields' (allowlist), 'omit' (denylist), 'unwind' (unwrap a nested field into top-level items) and 'flatten'.
7. For CSV specifically, control output with 'delimiter', 'bom' (emit a UTF-8 BOM so Excel reads accents correctly), 'skipHeaderRow', and 'attachment=true' to force a Content-Disposition download header.
8. Set 'desc=true' for newest-first ordering when you are incrementally exporting only recent items.
9. Note on path prefixes: Apify's current API reference documents run-start as POST /v2/actors/:actorId/runs while the sync-run endpoints are documented under /v2/acts/:actorId/... — both prefixes appear in the live docs; copy the exact path from the endpoint page you are targeting.
10. Docs: https://docs.apify.com/api/v2/dataset-items-get