Skip to content

User Guide

Verql is a fast, cross-platform desktop database client. You connect to your databases, browse their structure, write and run SQL, visualise results, and move data in and out — all from one app. It runs on macOS, Linux, and Windows, and it’s open source (MIT, by Arshad Shah).

Out of the box, Verql speaks PostgreSQL, MySQL, and SQLite, with MongoDB, Redis, and Snowflake added by bundled plugins. It includes a SQL editor with autocomplete, an interactive results grid, ER diagrams, charts, import/export tools, and a built-in AI assistant — and it can be extended further with plugins.

This guide is for people using Verql to work with databases. If you want to build a plugin or contribute to Verql itself, see the developer docs in the docs/ folder instead.

PageWhat it covers
InstallationInstalling on macOS, Linux, and Windows; verifying your download; auto-updates.
Connecting to a databaseCreating connection profiles, supported databases, SSH tunnels, and how your credentials are kept safe.
Running queriesThe SQL editor, autocomplete, the results grid, transactions, and the command palette.
Exploring your schemaThe schema browser, ER diagrams, table previews, the inspector, and charts.
Importing & exporting dataMoving data in and out as CSV, JSON, SQL, and JSON-Lines.
The AI assistantUsing OpenAI, Anthropic, or Ollama, tool-call approvals, and the built-in MCP server.
Managing pluginsWhat plugins are, how to enable/disable and install/uninstall them, and staying safe.
Keeping Verql updatedUpdate channels and how Homebrew upgrades work.
TroubleshootingCommon problems and how to report a bug.

When Verql opens you’ll see a three-panel workspace: a sidebar for your connections and schema, a central tabbed area for editors and table views, and context panels for details, charts, and the AI assistant.

A typical first session looks like:

  1. Install Verql for your platform.
  2. Create a connection to your database.
  3. Browse the schema to see what’s there.
  4. Open a query tab and run some SQL.
  5. Export the results — or ask the AI assistant for help.