DuckDB and VisiData are exceptionally capable. The csvlite interface removes the syntax.

Choose DuckDB for programmable analysis

DuckDB can query CSV and JSON directly, infer types, join files, use SQL, and feed Python and other analytical environments. For repeatable pipelines and complex relational work, it is the stronger tool. Its own documentation shows that a CSV can be queried with a filename or configured through read_csv.

Source: DuckDB data import overview.

Choose VisiData for a terminal-native workflow

VisiData describes itself as an interactive tabular-data multitool combining spreadsheet clarity, terminal efficiency and Python power. It offers keyboard-driven exploration, histograms, plots, session replay and many formats. Technical users who live in a terminal may find that workflow faster than any GUI.

Source: VisiData official site.

Choose csvlite for visual, discoverable exploration

With csvlite, text, regex, numeric and date filters appear as visible controls. Click a column to see its distribution and invalid values. Build grouped counts, sums, averages, minima, maxima and distinct lists without composing a query. Tabs, filters, sorts, widths and scroll position restore automatically.