Export
dump reads Google data from the Store. It never calls Google to fill a gap.
If Bing credentials exist, it also exports matching Bing data. Use --no-bing to skip Bing.
Every analytics row has site and search_type columns.
--format | Output |
|---|---|
parquet (default) | <site>/<search_type>/<table>.parquet |
csv, json, ndjson | <site>/<search_type>/<table>.<ext>, plus a per-row position |
sqlite | One gscdump.sqlite file. Dates are YYYY-MM-DD text |
duckdb | One gscdump.duckdb file |
Inspections, sitemaps, and Indexing API metadata go to <site>/<dataset>.<ext>, or to their own tables in a database file.
manifest.json lists every dataset with its row count, and the same coverage that sync --status reports.
--format sqlite needs Node.js 22.13 or later.
See export row limits for Google’s API and interface limits. To collect more dates before exporting, sync historical data.