phpMyAdmin
DBeaver
Sequel Pro
HeidiSQL
SQL Developer
MySQL Workbench
DataGrip
Adminer
Varan
DBeaver
DataGrip
TablePlus
Beekeeper Studio
DbVisualizer
DbGate
Navicat
Varan is a desktop SQL client for querying across different databases at once. Point it at MySQL, PostgreSQL, DuckDB, or spreadsheet files (CSV and Excel) and run a single SQL query โ including JOINs โ across all of them, without ETL pipelines or importing everything into one database first.
It runs on a local-first DuckDB engine, so queries execute on your own machine with read-only access to your sources: no admin rights, no server to set up, and nothing is written back to your databases.
Beyond cross-source SQL, Varan adds the tools data people always wish they had โ automatic anomaly detection when you open a table (duplicate keys, missing values, outliers, and orphaned foreign-key references), column-level lineage, git-style rollback of every change, and a Python surface on the same live tables you query in SQL.
Varan is a cross-platform desktop app for macOS, Windows, and Linux, currently in beta.
phpMyAdmin VaranVaran's answer:
Varan runs a single SQL query across different databases and files at once. Most SQL clients (DBeaver, DataGrip, TablePlus) connect to one database at a time; Varan lets you JOIN a MySQL table to a PostgreSQL table to a CSV in one statement, powered by a local-first DuckDB engine โ no ETL, no pipelines, no data copied into a warehouse first. On top of that it adds automatic anomaly detection, column-level lineage, and git-style rollback in the same workspace.
Varan's answer:
Data analysts, analytics engineers, and data-adjacent developers who regularly work across several databases and spreadsheets. People who can write SQL but aren't DBAs โ they have read-only access to production and don't want to spin up ETL pipelines or a warehouse just to answer an ad-hoc question spanning two systems.
Varan's answer:
Traditional SQL clients are single-database, so cross-source work means exporting and stitching data by hand. Federated engines like Trino or ClickHouse can query across sources but need infrastructure to stand up and manage. Varan sits in between: open a desktop app, point it at your sources (read-only, no admin), and query across all of them instantly โ with lineage, anomaly detection, and rollback built in. It's the convenience and workflow of a polished app on top of a serious engine, without the setup.
Varan's answer:
Varan started from a recurring frustration: whenever an answer needed data from two places โ a Postgres database, a MySQL host, and the inevitable sales_final.xlsx โ the options were to build an ETL pipeline, hand-write fragile pandas glue, or give up. We wanted to just point one tool at everything and write a normal JOIN. So we built Varan around a local-first DuckDB engine that queries across sources directly, and added the data-quality and versioning tools we always wished we had. It's currently in beta.
Varan's answer:
DuckDB (the local-first query engine) Electron + React (desktop application) Node.js (application/runtime layer) Python (bundled analysis surface) Java / Spring Boot (the licensing & sync hub backend) PostgreSQL (hub data store)
DBeaver - DBeaver - Universal Database Manager and SQL Client.
Sequel Pro - MySQL database management for Mac OS X
DataGrip - Tool for SQL and databases
HeidiSQL - HeidiSQL is a powerful and easy client for MySQL, MariaDB, Microsoft SQL Server and PostgreSQL. Open source and entirely free to use.
TablePlus - Easily edit database data and structure
SQL Developer - Oracle SQL Developer is a free, development environment that simplifies the management of Oracle Database in both traditional and Cloud deployments.
