CSV to SQL Converter — CREATE TABLE + INSERT, Type-Inferred

Real-world use cases

• Seed a dev database from a spreadsheet export

  You have a product catalog in Google Sheets and need it in your local Postgres for development. Export the sheet as CSV, paste it here, set the dialect to PostgreSQL, set the table name to `products`, and leave CREATE TABLE on. The tool infers `price` as DECIMAL, `in_stock` as BOOLEAN, `sku` as VARCHAR, and emits a CREATE TABLE plus a batched INSERT. Copy it, run `psql mydb < products.sql`, and your dev DB has real-shaped data in one paste instead of a hand-typed schema and 50 manual rows.

• Build a reproducible test fixture

  Your integration test needs 30 specific user rows. Keep them in a CSV under version control, convert with 'One INSERT per row' turned on, and commit the generated .sql alongside the test. Per-row INSERTs mean the diff shows exactly which row changed when someone edits the fixture, and a reviewer can comment out a single line to isolate a flaky case — far cleaner than a giant batched VALUES list.

• Migrate legacy data between databases

  You are moving a table out of an old MySQL box into SQLite for an offline app. Dump the source table to CSV, paste it, switch the dialect to SQLite, and the identifier quoting flips from backticks to double quotes and booleans render as 1/0. The generated CREATE TABLE gives you a starting schema you can tweak, and the INSERTs load straight into `sqlite3 app.db`.

• Quick-import a CSV someone emailed you

  A colleague sends a CSV of last month's signups and asks you to load it. Rather than fighting your client's import wizard and its type-guessing, paste the CSV here, turn CREATE TABLE off (the table already exists), keep batched INSERTs, set NULL-on-empty so blank cells do not become empty strings, and copy just the INSERTs into your SQL client. Done in under a minute, no upload, no temp file.

• Generate INSERTs for a documentation example

  You are writing a tutorial and need a realistic seed block readers can paste. Type a small CSV inline, generate the SQL, and paste the CREATE TABLE + INSERT into your docs. Because every value is escaped correctly — including the apostrophe in `O'Brien` — your example runs the first time instead of throwing a syntax error that derails a reader on step one.

Common pitfalls

• Pasting CSV without a header row and getting columns named col_1, col_2. This tool treats the first line as column names. If your export has no header, add one line of names first, or your CREATE TABLE will use generic identifiers that are hard to query.

• Forgetting that empty cells become NULL by default. If your column is NOT NULL or you genuinely want empty strings, turn off 'Empty cells as NULL' — otherwise the INSERT may violate a constraint or change the meaning of blank data.

• Trusting inferred types blindly for production schemas. A column of US ZIP codes like 02134 is all-digits, so it infers as INT and drops the leading zero. Review the CREATE TABLE and force VARCHAR for any code/ID column where leading zeros or formatting matter.

Privacy