Use File > Open File... (Ctrl+O) or drop a file on the window. Every file opens in its own tab. The type is detected from the extension: CSV/TSV/TXT, Excel .xlsx/.xlsm, JSON/.jsonl and XML. Anything else is named as unsupported instead of being shown as garbage.
Scroll, or type a row number into Go to row (Ctrl+G). Ctrl+Home goes to the first row, Ctrl+End to the last one. The grid keeps one page of 50,000 rows in memory and moves the page as you scroll, so the row count does not affect memory use.
Pick the worksheet in the top bar. Numbers and dates are displayed with the cell format stored in the workbook. View > First Row Is Header switches the header row when the automatic guess is wrong.
The structure tree on the left lists the arrays (JSON) or record nodes (XML) found in the file. Click one to make it the grid. View > Expand Nested Values turns nested objects and arrays into separate columns.
Search > Find... (Ctrl+F) scans the whole file and lists the hits; double-click a hit to jump to it. CSV, JSON and XML are scanned at byte speed; Excel has to decompress, so finding in a very large workbook takes longer.
Ctrl+A selects the page, clicking a column header selects the column, clicking a row number selects the row (Ctrl+C copies as TSV). File > Export View as CSV... writes what the grid currently holds, header row included, as CSV or TSV.
BigTextFileViewer — the same idea for huge text and log files
DataFileConverter — convert a big file to another format
DataFileSplitter — split a big data file into smaller ones
Withdata blog — how-to articles about large files and data conversion