Beekeeper Studio connects to SQL databases, browses schemas and tables, edits rows and runs SQL in query tabs. A connection gives the client access under the supplied database account. It does not copy the whole database into a local document. Query results remain a view of returned rows until the user exports them or applies an edit to the database.
The result grid has a row boundary
The normal result view and its download action limit a query to twenty thousand records by default. A large query can therefore look complete in the grid even when the database returned more matching rows. Sorting or inspecting only the visible result does not reveal the missing tail. The selected export route determines whether that cap remains.
Download full results removes that cap by streaming the complete query result to disk. This path narrows the formats to CSV and JSON. The ordinary capped result can export CSV, Excel, JSON, Markdown or an Excel-friendly TSV file, so choosing full results trades format choices for completeness.
Table export streams separate files
A table export reads rows from the database and can write CSV, SQL INSERT, JSON or JSON Lines output. When the user selects several tables, Beekeeper Studio writes each table to its own file. It does not create one relational archive that automatically preserves foreign-key order or restores an entire schema.
Streaming avoids holding every row in the interface, but a full table can take a long time. The database account, server load and network connection remain part of the operation. Interrupting the stream can leave a partial output file, so the saved size or final record needs verification before the export replaces another backup.
Copying selected rows is not a full export
Selected rows can be copied as TSV, JSON, Markdown or SQL INSERT statements. This is useful for a small sample or migration fragment. The selection follows the rows currently present in the grid and cannot recover rows hidden by the result limit.
SQL INSERT output represents data statements rather than the complete table definition. It may need a compatible destination schema and appropriate ordering. Markdown provides a readable table but loses database types and constraints. The format should match whether the next step is inspection, spreadsheet work or execution against another database.
Connection scope controls what editing can change
Saved connections and query tabs are separate from table data. Editing a row sends a database change under the active connection; closing the tab does not act like discarding an unsaved text document after that change reaches the server. Before running an UPDATE or DELETE statement, checking the active database and schema prevents a correct query from changing the wrong database environment.
