Data Studio

Bar, grouped bar, line, area, pie, donut, scatter, histogram, heatmap, and pivot table.

Yes — full control over colours, axis labels, title, and legend.

No. Everything runs locally in your browser.

Process files up to the browser memory limit.

Yes — export as PNG, SVG, or an interactive HTML file.

Yes. Upload multiple files using the multi-file drop zone. Each file becomes its own table. Write a standard SQL JOIN across the table names, and DuckDB executes the join in-memory. For example: SELECT a.name, b.revenue FROM customers a JOIN sales b ON a.id = b.customer_id. This is a common workflow for analysts combining a CRM export with a billing export.

Data Studio uses DuckDB's SQL dialect, which is largely compatible with PostgreSQL syntax. It supports CTEs (WITH clauses), window functions (ROW_NUMBER, RANK, LAG, LEAD), ARRAY functions, JSON functions, PIVOT, and UNNEST. DuckDB also supports reading Parquet and JSON natively, in addition to CSV.

Pyodide ships with a large set of pre-compiled scientific Python packages: Pandas, NumPy, Matplotlib, SciPy, scikit-learn, and many others. Pure-Python packages from PyPI can also be installed at runtime using micropip.install('package-name'). C-extension packages that are not pre-compiled for WASM are not supported.