Panshi

Generate SQL from a CSV and run it in your browser

Upload a CSV, ask a question in plain English, and get both the SQL and the actual answer — computed locally, your data never leaves your device.

You have a spreadsheet and a question — "average revenue per category", "top 10 customers", "rows where status is overdue" — but writing the SQL or wrangling a pivot table is friction. This removes it.

Upload the CSV (it’s parsed in your browser; the file is never uploaded), type the question, and the tool generates the SQL and runs it locally against your data, returning the answer as a table you can download. Only your question and the inferred column names go to the model.

It’s ideal for quick analysis without spinning up a database — and you can copy the generated SQL to reuse elsewhere.

The tool for this

🗄️Text to SQL

Turn plain English into correct SQL for Postgres, MySQL, SQLite, BigQuery or Snowflake — schema-aware.

Try Text to SQL →

Frequently asked questions

How do I query a CSV with SQL without a database? +

Upload the CSV here; the tool generates SQL from your question and runs it in your browser, returning a downloadable result.

Does my CSV data get uploaded? +

No — the file is parsed locally and the query runs in your browser. Only your question and inferred column names are sent.

Can I reuse the generated SQL? +

Yes — copy it and run it against your real database; it’s standard SQL.

What kinds of questions work? +

Aggregations, filters, top-N, grouping and joins-within-the-file all work, e.g. "average amount per category".

Related tools

Browse the full tools directory, or see all Panshi services.