TinyPDF.in

Database Playgrounds

Practise SQL and MongoDB queries without installing a database. Everything runs in your browser.

PostgreSQL PlaygroundJSONB, window functions, CTEs, PL/pgSQL. Import CSV, JSON or .sql.SQLPostgreSQLSQLite PlaygroundOpen .sqlite and .db files, query them, export the database.SQLSQLiteDuckDB PlaygroundQuery CSV, JSON and Parquet files. PIVOT, QUALIFY, SUMMARIZE.SQLDuckDB SimulatorMongo Query Playgroundfind, aggregate, $lookup, $group and updates on sample collections.Mongo shellTinyPDF.inPostgreSQL Query Performance LabCompare B-tree, hash, BRIN and GIN indexes on the same query, or pgvector HNSW vs IVFFlat for vector search with recall.SQLEXPLAIN ANALYZEpgvectorpgvector PlaygroundVector search and RAG queries on document chunks with embeddings, in real PostgreSQL.SQLPostgreSQL + pgvector

For learning and small datasets. These run inside your browser, so Java, Python and other programs cannot connect to them.

About these playgrounds

Each playground runs a database engine inside your browser tab and starts with a small sample database, so you can write your first query straight away. You can also import your own CSV, JSON, Parquet, .sql or .sqlite files.

Use them to learn SQL or MongoDB syntax, test a query before running it on a real server, or look inside a small data file. They are not hosted databases: there is no connection string for applications, and data is cleared when you reload.

Questions

Which playground should I use?

PostgreSQL for learning standard SQL and Postgres features, SQLite for opening .sqlite files, DuckDB for analysing CSV or Parquet files, and Mongo for practising MongoDB queries.

Can I connect from Java, Python or another program?

No. The databases run inside your browser, with no server or connection string, so applications and drivers cannot connect.

Is my data uploaded?

No. Queries and imported files stay on your device.

Do I need an account?

No. Open a playground and start writing queries.